Determining the Proper Oil for Your Engine - Email from Lake Speed Jr

[rstcso]

I changed the oil at 494 miles to Amsoil Signature Series 0w-20 in the new Honda Pilot and sent UOA and VOA samples to Speediagnostix. I asked Lake if my plan to change it again after another 4k miles made sense, and whether I should increase the viscosity to 0w-30. This is his comprehensive reply, which he has given permission to post here. Based on his recommendation in Step 1, I should have continued using the OEM oil. Too late now, which he acknowledged.

"Regarding your plan, here’s the proper framework for determining the proper oil for your engine.

Step 1 - Utilize the OEM recommended oil and do two early oil changes during the break-in process (500 to 1,000 miles and again between 3,000 and 4,000 miles). It is ok that you switched to Amsoil Signature Series already. Just stick with it to establish the baseline.

Step 2 - Take used oil samples at each oil change to establish the trend analysis.

Step 3 - Go 5,000 miles on the third oil change and take a used oil sample. If the wear rate per 1,000 miles is below 5 ppm, you are good. If the wear rate is between 5 ppm and 10 ppm per 1,000 miles, go another 5,000 miles on the OEM recommended oil and resample. If the wear rate is still greater than 5 ppm per 1,000 miles, then move to step 4.

Step 4 - Since the OEM recommended oil and viscosity have not produced a wear rate per 1,000 miles lower than 5 ppm, go up to next viscosity grade in the OEM oil. Go 4,000 to 5,000 miles on that oil and then take another sample. See if the change in viscosity drops the wear rate per 1,000 miles below 5 ppm. If it does, you are good. If it does not, then move to Step 5.

Step 5 - Since the change in viscosity did not get the wear rate per 1,000 miles below 5 ppm, try a different brand of oil in the same viscosity grade of whichever oil had the lowest wear rate per 1,000 miles. You will need to use it for 3,000 to 4,000 miles to flush the OEM oil out of the system before going 5,000 miles on the new oil to take another sample. See if the non-OEM oil lowers the wear rate per 1,000 miles to 5 ppm or less. If it does, you are good. You can then use the oil analysis results to fine tune the oil change interval.

If the wear rate per 1,000 miles is still above 5 ppm, try the next higher viscosity oil of that same brand to see if that lowers the wear rate per 1,000 miles. Finding the best oil for an engine is an iterative process, but the data from the samples (viscosity, additive depletion, wear rate) will paint a picture that guides you in the right direction.

The key is to make decisions based on facts instead of fear. Use the science instead of speculating.

Thanks,
Lake Speed Jr."

 

[tojo7089]

Forgive my ignorance.......5ppm/1000 miles of what?

 

[rstcso]

Forgive my ignorance.......5ppm/1000 miles of what?

He talks about wear rate. The bottom row shows the wear rater per 1000 miles. Miles on oil/1000. Shown are the results of my initial oil change at 494 miles.

It's important to know what metals may be part of an oil's additive package, which needs to be subtracted out of the metals shown. The second screen shot under Test Results shows a VOA of Ravenol VSE 0w-20 having 4ppm Tin, which Lake says is due to the Molybdenum. If I continued to use this oil, any higher Tin numbers may be due to engine wear, not the additive package.

 

[cptbarkey]

His generic advice doesnt state the sample rate (via universal averages) on the same vehicle and engine. This is the one benefit you would gain from Blackstones UOA services, alibet with inaccurate viscosity and fuel dilution. For example, its well known that Hemi and Pentastar engines shed iron and copper normally, far above his generic indicators.

I certainly wouldnt increase viscosity on his advice.

 

[Patrick0525]

I guess at 1.40 ppm/1k, I am good to go at a 9.5K OCI.

HPL euro 5w-30

 

[rstcso]

I guess at 1.40 ppm/1k, I am good to go at a 9.5K OCI.

HPL euro 5w-30

And if your owner's manual called for 10k OCIs, the UOA would have told you otherwise, by 500 miles!

 

[ARB1977]

I ran SS 0W20 up until recently and switched to 0W and 5W30. I’m going to Amsoil euro 0W30 next. No uoa, follow the olm and carry on.

 

[Mechanician]

Life's too short .....

 

[rstcso]

I ran SS 0W20 up until recently and switched to 0W and 5W30. I’m going to Amsoil euro 0W30 next. No uoa, follow the olm and carry on.

I had 185k miles on my 2002 Ram 1500 4.7L before selling it to a neighbor. Always ran Amsoil. Still running without any engine failures over 200k miles. Imagine the longevity had I been doing UOAs!

 

[rstcso]

Life's too short .....

That it is. If I follow his plan, it will be at least three years before we put on enough miles to get to the last step.

 

[Klutch9]

That it is. If I follow his plan, it will be at least three years before we put on enough miles to get to the last step.

I think his methodology is sound. But yes, it will take awhile to establish that baseline.

Thankfully for me, there is a guy on FB who has a Kia Carnival like mine and has done numerous UOAs with the same oil, same grade. His wear metals are super low even with varied OCI length and time. I plan on using the EP version of the oil he was using, and I'm expecting similar results thanks to his data.

 

[Mechanician]

That it is. If I follow his plan, it will be at least three years before we put on enough miles to get to the last step.

If you had a race team, or your livelihood depended on the engine's livelihood - this level of scrutiny would definitely be worth the effort.
Otherwise admitting OCD (Oil Compulsive Disorder) may be a good idea.

 

[dnewton3]

LSJr's approach to wear metal accumulations are, well, misinformed at the very least.
He is, IMO, making good money off giving average-at-best advice.
He makes three very important mistakes in his approach:
- he combines wear metals to get a "total" per 1k miles; as if those metals can be lumped together to understand wear traits ... (SMH)
- he completely ignores the topics of macro and micro statistical analysis
- he completely ignores the topics of understanding process variation

UOAs are great tools, but they have to be understood as to what they are and are not good at. And more importantly, once you have the data, how to accurately and properly process that information.

 

[dnewton3]

I think his methodology is sound.

NO! It's not sound by a long shot.
Much of the problem is rooted in the fact that chemists and tribologists aren't necessarily good at statistical analysis; they don't know the proper and improper ways to crunch data. LSJrs may be a trained oil geek, but he's clearly a noob when it comes to processing data.


Want to understand how to properly use UOAs?
Read this:
https://bobistheoilguy.com/used-oil-analysis-how-to-decide-what-is-normal/

 

[FranklinJL]

LSJr's approach to wear metal accumulations are, well, misinformed at the very least.
He is, IMO, making good money off giving average-at-best advice.
He makes three very important mistakes in his approach:
- he combines wear metals to get a "total" per 1k miles; as if those metals can be lumped together to understand wear traits ... (SMH)
- he completely ignores the topics of macro and micro statistical analysis
- he completely ignores the topics of understanding process variation

UOAs are great tools, but they have to be understood as to what they are and are not good at. And more importantly, once you have the data, how to accurately and properly process that information.

Im sorry, not trying to defend the guy, but he was employed by actual race teams.......I think he knows what he's doing. He's a science based guy, vs speculation......how could you call the guy a noob if he helped develop oils for race teams? I mean the internet is something else..

 

[The Critic]

Im sorry, not trying to defend the guy, but he was employed by actual race teams.......I think he knows what he's doing. He's a science based guy, vs speculation......how could you call the guy a noob if he helped develop oils for race teams? I mean the internet is something else..

Marketplace reputation is a valid yardstick, and to be fair, the guy was probably giving “watered down advice” to the OP to make it more understandable/relatable to the average car enthusiast.

However, if you were to scientifically dissect his recommendations, they are arguably flawed.

 

[gfh77665]

NO! It's not sound by a long shot.
LSJrs may be a trained oil geek, but he's clearly a noob when it comes to processing data.

I agree and have said the same before. I don't bash LSJ. He has good content, and he is an intelligent guy. The concern I have with him is his data interpretation. He means well (and he likes clicks of course) but his data interpretation and conclusions needs to be taken with a grain of salt. For example, I have seen him take just one data point, and run off with it. Scientific methodology just does not work that way.

 

[Klutch9]

NO! It's not sound by a long shot.
Much of the problem is rooted in the fact that chemists and tribologists aren't necessarily good at statistical analysis; they don't know the proper and improper ways to crunch data. LSJrs may be a trained oil geek, but he's clearly a noob when it comes to processing data.


Want to understand how to properly use UOAs?
Read this:
https://bobistheoilguy.com/used-oil-analysis-how-to-decide-what-is-normal/

I guess what I was getting at is that his suggestion to sample the same oil numerous times might provide someone with a consistent set of micro data. Maybe; depends on their ability to maintain specific driving habits and sample consistently. It’s better than nothing and might help track break in, if nothing else. What someone infers from it is up to them.

Let’s say I do a number of UOAs that show I pretty consistently have 1.5ppm/1,000 miles of wear metals. Then I get a sample that’s much higher, say 4ppm, with one or two metals that stand out. This could indicate that something is amiss at the very least, if all other variables are equal.

Of course, I defer to those who know more than I do about this. I’ve never done a UOA personally.

 

[FranklinJL]

I guess what I was getting at is that his suggestion to sample the same oil numerous times might provide someone with a consistent set of micro data. Maybe; depends on their ability to maintain specific driving habits and sample consistently. It’s better than nothing and might help track break in, if nothing else. What someone infers from it is up to them.

Of course, I defer to those who know more than I do about this. I’ve never done a UOA personally.

Im still blown away that people will fully question a guy that was employed by a race team for years on this very subject. Im like 100% sure he knows more than any of us about motor oil. It's like LSPjr vs a guy that works at Jiffy Lube.....what a forum lol

 

[TomYoung]

If you had a race team, or your livelihood depended on the engine's livelihood - this level of scrutiny would definitely be worth the effort.
Otherwise admitting OCD (Oil Compulsive Disorder) may be a good idea.

I admit it.