2010 FX4 | M1 0W-20 AFE SN | 5.4L | 15,372 miles

Status
Not open for further replies.
Completely agree; a PC is a much better method to judge filtration efficiency than is a UOA.

But as you noted, and I've said countless times, once a minimum efficiency is met to satisfy the engine, anything "more" (better, etc) really does not manifest into tangible results in true wear data.

There are three contributors to the wear issue, and wear is what we should be most concerned about.
1) add pack
2) filtration
3) OCI duration

Those work in concert to establish low wear. But they are not linear in their response, and they are not mutually exclusive. From everything I've seen, the wear is most greatly affected by the barrier layers from the oxidation of oil. Run it longer, the wear goes to near-zero.

It's the whole topic of inputs versus outputs. Hence, my constant harping on the ROI (be it lubes or filters). At some point, good enough is good enough, and more ain't necessarily better, unless you greatly shift the operational terms (supe- duper cold, silly-long OCIs, etc). This is, of course, predicated on a clean running engine that has no neglect in its history.
 
Last edited:
Originally Posted By: Jim Allen
It's beginning to look to me that the oil will do more in that regard than high-end filtration... at least when the filtration is above a minimum standard of around 35-40 microns absolute. This UOA adds another grain of truth to that. These Mobil 1 UOAs appear to be "better" that those on the M5K.

I agree with your thoughts--since the filters have been FL-820s to this point in time and if we assume the quality control on them is more or less the same so thus the filters are more or less the same making them a constant in the equation. The driving style has remained more or less the same from day zero so it is also somewhat of a constant. What has changed 4 times in the equation is the type of oil and the duration upon which each has been used. I would need to run M1 more times to equal the MS5K runs, but from where I am standing the M1 performed "better" than MS5K and the PU. I never ran PU out to 10K or beyond so in its defense perhaps it would produce better wear numbers than it did at 7.5K, but the insolubles speak for themselves for both MS5K and PU.
 
I would caution both of you against making such a quick determination ...

You would need to do a LOT more testing to truly call one "better" than another. You'd have to establish the running ranges and look for trends. Two OCIs does not make for good micro analysis. The UOAs can be compared and contrasted to macro data; in that regard, all your samples are well within normal sigma deviation.

And NONE of your samples ever remotely approached a true point of end-of-lifecycle based upon condemation limits. You did a great job of stretching out the OCIs, but not to a point where the lubes or the wear data indicated that an actual change was due; you changed oil based upon desire, not science. The only metal that varies is the Fe, and it's nowhere near any point of condemnation in ppm. All other metals are noise. The TBN/TAN has never shown to be a danger, nor vis, nor FP, etc.

Simply put, you are certainly entitled to your opinion, but I would point out that the M1 is not "better" because you've not run a true trial to establish the real statistical performance of any lube in a micro sense; you have only run lubes against the average of mass market equations.

Yes, I am nit-picking here. But this is the very point I try to emphasize in my "normalcy" article. I do not disagree that all your lubes did well. But that is only in view of macro data; you have no micro data worthy of proclaiming any one of them a successor, as you have way too few samples.
 
Last edited:
Originally Posted By: dnewton3
Two OCIs does not make for good micro analysis. The UOAs can be compared and contrasted to macro data; in that regard, all your samples are well within normal sigma deviation.
I have already acknowledged that in my previous post. I will NEVER have the amount of samples required to meet the criteria that you state is needed for such an analysis--because at some point in the near future I will likely stop UOA'ing the oil.

Originally Posted By: dnewton3
And NONE of your samples ever remotely approached a true point of end-of-lifecycle based upon condemation limits. You did a great job of stretching out the OCIs, but not to a point where the lubes or the wear data indicated that an actual change was due; you changed oil based upon desire, not science. The only metal that varies is the Fe, and it's nowhere near any point of condemnation in ppm. All other metals are noise. The TBN/TAN has never shown to be a danger, nor vis, nor FP, etc.
Let us not be penny wise and pound foolish here Dave! I will never run the oil to the point of Fe condemnation because I am certain the TBN/TAN will be low/high enough to warrant a change long before reaching that number (whatever it may be). I had never planned to use the physical attribute to condemn the oil (Fe), it was always going to be the chemical one (i.e. TBN/TAN levels). While I have a bit more to go to hit the chemical limits, I am rapidly approaching them. My "experiment" in all of this was to prove the that oils are underused and are changed far too early, in the right context conventionals can perform as well as a synthetic, and that xW-20 is not a monster that wears engines faster than a heavier oil. I think I have proven the former two and we will see about that latter. However, (I understand that it is only 2 UOAs), the M1 produced less Fe per mile than PU and dramatically less than MS5K and the engine runs smoother on M1. Regardless of the lack of comparative UOAs, this means something.

Originally Posted By: dnewton3
Simply put, you are certainly entitled to your opinion, but I would point out that the M1 is not "better" because you've not run a true trial to establish the real statistical performance of any lube in a micro sense; you have only run lubes against the average of mass market equations. Yes, I am nit-picking here. But this is the very point I try to emphasize in my "normalcy" article. I do not disagree that all your lubes did well. But that is only in view of macro data; you have no micro data worthy of proclaiming any one of them a successor, as you have way too few samples.

Perhaps to the point of not being realistic. Dave, you must realize he average joe would never create a UOA data stream like I have and even I have a "limit". If I recall, you stated to have enough samples for a true analysis would number in the hundreds (correct me if my memory fails me). This would equate to millions of miles and countless amounts of money and is something that is never going to happen (at least for me). At the end of the day, whatever number of samples I do will be what it is, but it is my belief that I am LISTENING to the UOA data and as more M1 samples come in a better comparison can be made, but there can be no argument that M1 produced less Fe than either of the other oils.
 
Originally Posted By: dnewton3
The insolubles count in the Blackstone UOAs is a characterization of the darkness of the oil after treated and then processed in a centrifuge and rated against a visual coloration standard. There is no real way to put any accurate measure of the insol count to a particular efficiency of a filter. Soot and oxidation are what contribute chiefly to the insol count. That being known, there is no ability to attribute the darkness to the size of particles.

I have seen average filers and great filters all turn in insol counts anywhere from .2 to .4; there is very little disparity between filters in this regard.

After reviewing over 10,000 UOAs, and a lot of them being Blackstone, I can tell you that there is no ability to show any correlation between wear and filter selection in the typical UOA.



I got the idea about filters and insolubles from Blackstone themselves on a recent UAO I had on my truck.

This engine is wearing like a Toyota, which is good since they tend to wear very well. Universal
averages show typical wear levels for this type of engine after about 5,300 miles on the oil. This oil wasn't
run too much longer than that so it was nice to see all wear matching up nicely to our averages. The oil itself
held up nicely too. The TBN was 3.0 showing plenty of active additive left; less than 1.0 is considered too
low. Low insolubles show the oil filter was working well. Try going up to 7,500 miles on the next oil and
check back to see how everything held up.
 
Last edited:
2010_FX4

Sorry if it came off harsh; did not mean it to be.

But I thought you and others understood the difference here, but you and Jim fell right back into the old trap. You called the M1 "better" only after two UOAs.

It takes at least 30 samples to establish a reasonable mean and standard deviation node. So, for each lube you try, you'd have to run 30 UOAs. And when you vary the OCI duration, you're changing inputs. I have always said there is not a BITOGer here I know of that has the time/money to really call something "better" from a truely scientific perspective. I include myself and you and everyone else in that. But what we can do is compare/contrast our individual UOAs to known averages, and then see how they stack up, given our choices of inputs. Your M1 runs did not even spike out of the first node of sigma deviation. While you may perceive that the M1 was better than the MS5K, it really was not. All your lube reports are totally normal for the exposure.

As for the TBN/TAN relationship, your condemation points are fine, but we've all discussed before the fact that TBN/TAN are only predictors of possibilities; in and of themselves they mean nothing. Only if that TBN/TAN affects wear, would you then be able to claim that the relationship had merit. Until that time, it's moot. I've shown that TAN crossing over TBN, even on order of 2x, means nothing to the wear data. If you choose to use TBN/TAN as condemnation, that's fine. But you're doing so arbitrarily.

What I am impressed with is that you stick to one lube for a reasonable amount of time, to establish that it's "normal" in it's performance. I applaud you for that. Both publically and privately, I've praised you for your approach. But I cannot agree with this new conclusion you have come up with; you are misinterpreting the data.

But you're falling prey to the wrong convictions here. You have no ability to safely and accurately say any lube you have tried is better than another lube you've tried. You have zero data to support that conclusion. You have anecdotal evidence that your lubes all perform to a very satifying standard, where the engine most likely outlast your use for it, regardless of what lube you use.

When I ran my longer OCI trials, I never claimed that the ST lube was "better" than any other in terms wear. To the contrary, I called out the fact that the performance of it was "normal", showing that it was on par with all other options, and that the longer OCIs had zero negative effects. Yours are the same; you are proving that longer OCIs are safe and practical. You are showing that the syns and dinos all perform (essentially) the same.

EVERYTHING we do should be targeted as a question of how it effects wear. If it does NOT cause effects in wear ranges and trends, then it is either ambiguous, or not at a magnitude that yet has merit. If wear were fine, would it matter if vis was too high or low? If wear were fine, would it matter if you added one packet of KoolAid to the OCI? If wear were fine, would it matter if you ran goat urine rather that oil? EVERYTHING we do should be studied in refernce to wear. You can look at other things as predictors, but those are only inputs. The output (wear data) is what matters. Vis is low? Tan is high? FP low? If the wear is not shifted, it means nothing other than the equipment is not yet affected. This topic reminds me of the recent movie "Moneyball" ... it does not matter if the player has a nice swing, or is a leftie, or etc. What matters is pushing up the score, and that is only done by getting on base. The same is here for engines; wear is what takes equipment from use, period. If your variable is not effecting wear, it is either unable to do so, or not in such quantity as to make a difference. Period.

Again - sorry to be blunt, but the point needs to be understood. UOAs are great tools, but are often used for improper conclusions. For you and Jim to call M1 "better" is fine for you to opine, but the facts are not supportive of your position. All of your UOAs are "normal", and nothing more. You are proving that syns and dinos in your application are essentially the same, and that they can be safely run for longer durations than the OEM prescribes. Nothing more; nothing less.


If I have offended you, then I owe you, and offer, this appology. I am sorry.
 
Last edited:
2010_FX4
"M1 produced less Fe than either of the other oils"

Correlation does not imply causation - it's possible the engine is simply more broken in now and wear is decreasing...period. Lubricant type may or may not have been a factor.
 
Originally Posted By: dnewton3
2010_FX4

Sorry if it came off harsh; did not mean it to be.


If I have offended you, then I owe you, and offer, this appology. I am sorry.



Harsh? Apology? I am from DC. "Don't apologize. Don't explain".
 
Originally Posted By: bigt61
2010_FX4
"M1 produced less Fe than either of the other oils"

Correlation does not imply causation - it's possible the engine is simply more broken in now and wear is decreasing...period. Lubricant type may or may not have been a factor.

More broken-in now at 122K versus...what 105K or 95K? Time will tell if that is true.
 
Originally Posted By: dnewton3
Sorry if it came off harsh; did not mean it to be. If I have offended you, then I owe you, and offer, this appology. I am sorry.

Dave; it is all good--no way each of us will always see eye to eye and no apology needed.

My definition of "better" is that M1 produced less wear metal per mile and less insolubles than the other oils. The issue that I have with your stance is the MS5K data was all too easily accepted without so much as a flicker but somehow the M1 data is disputed.

After 3 additional UOAs I will have the equivalent of the MS5K runs and although there will still only be 5 each there is a small amount of trending that can be compared. So we will have to see where the data takes us. I am by no means a statistician and would not dive into the details as you may, so I retract my "better" statement, but stand by my opinion that M1's performance outpaced MS5K and PU. Time will tell if my theory and opinion hold sway.
cheers3.gif
 
What I do believe, and your data may be able to prove out, with some adjustment to the methodology, is that (as most of us suspect) the syn will beging to outperform the dino at some distance. The questions become, at what distance, and to what measure, in regard to ROI? To know if something is "better", you have to have a baseline of measurement. Since you're not doing micro analysis, you have to turn to macro data and view "normalcy". You cannot attempt some hybrid form of blended analysis; it can't be done with any expectation of trustworthy results.

Pay attention to wear; that is what matters. Don't sweat the sideline items; they are only predictors. At 10k mile intervals, it's not like vis or FP or insols or TAN is going to lead to imminent engine demise ...

I would ask that you consider restraining the inputs to less variation, and ONLY vary the lubes. You have spent a great deal of time and money (as have I) proving that longer OCIs are perfectly safe. You inched up there, whereas I plundered in; but I think we can agree that 15k miles in not a real risk as most would believe. That in mind, I'd suggest this:
Pick and OCI you're comfortable with, and then manipulate only one variable (Oil); say 10k miles for instance. Then run three back-to-back OCIs with a syn, then three more with a dino. While you cannot compare/contrast them directly, you can then see if they are "normal" in response; I suspect they certainly will be. Once that is confirmed or denied, then you can walk out the OCI duration a bit more, and look for the disparity where one lube might not respond "normaly". THAT will be the point at which you can discern one lube is not able to keep up with mass-market averages and deviation. Then, you need to do the same for the alternative lube. Only then, once you know the limit of each lube, can you judge how much distance it would take for one lube to pay back any increase in cost structure.

In short, macro-normalcy does not allow you to judge one lube against another; it only allows comparison/contrast against mass-market response. But you can see at what distance each lube would be pushed outside of "normal" performance, and then see the relative disparity (delta) and how it figures into ROI. If it takes 12k miles for the dino lube to got outside "normal", and it takes 20k miles for the syn to go outside "normal", then the syn would need to cost less than that 67% differential just to break even.

Or, you could run it the other direction. Look at the cost of both, and see if the syn can provide the distance equivilant in ROI; if it costs 2x as much, did it run 2x further before it went outside "normal" response?

I hope everyone sees the difference I'm trying to convey here:
Micro analysis is VERY time consuming and expensive. It will allow you to DIRECTLY compare/contrast two lubes against each other. But unless you ran full trials, you don't have to true averages and standard devation. (It's a geek math thing; too lengthy to explain here). This allows a true statement of "better" (once you define what better means to you).
Macro data allows you to compare/contrast your results to mass market response, and you have to see how your choices fare against the "normal" data set, and then see what it takes to push each alternative outside normal response. Here, you are doing INDIRECT comparison/contrast, and making an inference based upon a shift outside "normalcy". "Better" in this case only infers that one alternative held on to "normal" for a longer period.

Make sense?
 
Last edited:
Originally Posted By: 2010_FX4
Originally Posted By: dnewton3
Sorry if it came off harsh; did not mean it to be. If I have offended you, then I owe you, and offer, this appology. I am sorry.

Dave; it is all good--no way each of us will always see eye to eye and no apology needed.

My definition of "better" is that M1 produced less wear metal per mile and less insolubles than the other oils. The issue that I have with your stance is the MS5K data was all too easily accepted without so much as a flicker but somehow the M1 data is disputed.

After 3 additional UOAs I will have the equivalent of the MS5K runs and although there will still only be 5 each there is a small amount of trending that can be compared. So we will have to see where the data takes us. I am by no means a statistician and would not dive into the details as you may, so I retract my "better" statement, but stand by my opinion that M1's performance outpaced MS5K and PU. Time will tell if my theory and opinion hold sway.
cheers3.gif




Less wear metals per mile is more evidence that wear metals lessen as the miles rack up because of the anti-wear layer put on by the oil.
Just spitballing,although it does make sense.

And isn't a mod motor just getting broken in at 100000 miles........
 
Originally Posted By: dnewton3
2010_FX4

Sorry if it came off harsh; did not mean it to be.

But I thought you and others understood the difference here, but you and Jim fell right back into the old trap. You called the M1 "better" only after two UOAs.

It takes at least 30 samples to establish a reasonable mean and standard deviation node. So, for each lube you try, you'd have to run 30 UOAs. And when you vary the OCI duration, you're changing inputs. I have always said there is not a BITOGer here I know of that has the time/money to really call something "better" from a truely scientific perspective. I include myself and you and everyone else in that. But what we can do is compare/contrast our individual UOAs to known averages, and then see how they stack up, given our choices of inputs. Your M1 runs did not even spike out of the first node of sigma deviation. While you may perceive that the M1 was better than the MS5K, it really was not. All your lube reports are totally normal for the exposure.

As for the TBN/TAN relationship, your condemation points are fine, but we've all discussed before the fact that TBN/TAN are only predictors of possibilities; in and of themselves they mean nothing. Only if that TBN/TAN affects wear, would you then be able to claim that the relationship had merit. Until that time, it's moot. I've shown that TAN crossing over TBN, even on order of 2x, means nothing to the wear data. If you choose to use TBN/TAN as condemnation, that's fine. But you're doing so arbitrarily.

What I am impressed with is that you stick to one lube for a reasonable amount of time, to establish that it's "normal" in it's performance. I applaud you for that. Both publically and privately, I've praised you for your approach. But I cannot agree with this new conclusion you have come up with; you are misinterpreting the data.

But you're falling prey to the wrong convictions here. You have no ability to safely and accurately say any lube you have tried is better than another lube you've tried. You have zero data to support that conclusion. You have anecdotal evidence that your lubes all perform to a very satifying standard, where the engine most likely outlast your use for it, regardless of what lube you use.

When I ran my longer OCI trials, I never claimed that the ST lube was "better" than any other in terms wear. To the contrary, I called out the fact that the performance of it was "normal", showing that it was on par with all other options, and that the longer OCIs had zero negative effects. Yours are the same; you are proving that longer OCIs are safe and practical. You are showing that the syns and dinos all perform (essentially) the same.

EVERYTHING we do should be targeted as a question of how it effects wear. If it does NOT cause effects in wear ranges and trends, then it is either ambiguous, or not at a magnitude that yet has merit. If wear were fine, would it matter if vis was too high or low? If wear were fine, would it matter if you added one packet of KoolAid to the OCI? If wear were fine, would it matter if you ran goat urine rather that oil? EVERYTHING we do should be studied in refernce to wear. You can look at other things as predictors, but those are only inputs. The output (wear data) is what matters. Vis is low? Tan is high? FP low? If the wear is not shifted, it means nothing other than the equipment is not yet affected. This topic reminds me of the recent movie "Moneyball" ... it does not matter if the player has a nice swing, or is a leftie, or etc. What matters is pushing up the score, and that is only done by getting on base. The same is here for engines; wear is what takes equipment from use, period. If your variable is not effecting wear, it is either unable to do so, or not in such quantity as to make a difference. Period.

Again - sorry to be blunt, but the point needs to be understood. UOAs are great tools, but are often used for improper conclusions. For you and Jim to call M1 "better" is fine for you to opine, but the facts are not supportive of your position. All of your UOAs are "normal", and nothing more. You are proving that syns and dinos in your application are essentially the same, and that they can be safely run for longer durations than the OEM prescribes. Nothing more; nothing less.


If I have offended you, then I owe you, and offer, this appology. I am sorry.


MONKEY NIPPLE'S
39.gif
 
Originally Posted By: dnewton3
I would ask that you consider restraining the inputs to less variation, and ONLY vary the lubes.

Less variation? Do you mean miles per OCI or? Other than the oils, only two things have changed during this entire 122K run of UOAs (19 of them thus far)--the duration of OCI and (on this last run) the oil filter brand. Aside from that all other things have remained as constant as feasible (unless I am missing something). I have to think I had very few variations in the grand scheme.

As for further experimenting, I will see what the next two runs of AFE reveals and decide from there. I am nearly to the end of the experimentation and will be settling into a routine--M1 may be the final choice or it may not. I have quite a stash of M1 that I inherited from my Dad, but that can be used in my new Explorer, so nothing is set in stone just yet.

Originally Posted By: dnewton3
Pay attention to wear; that is what matters.

Are we talking about the wear metals in the UOAs? I just want to be clear I understand to what you refer here.
 
Sorry I wasn't clear on my suggestion.

Yes - I would recommend one OCI limit at first; pick 10k miles if you wish. Run all your UOAs to that limit. By eliminating the OCI duration as a variable, you take one input and make it a constant. I fully realize you've been slowly and methodically extending your OCIs and I applaud that, but you've proven that 15k miles is clearly safe in your situation. That in mind, choose an OCI (10k miles?) and make that a steady input. So ONLY the oil is the studied variable. I didn't mean to infer you have been willy-nilly in your approach; to the contrary I very much applaud your steady state mentality. I only am trying to offer a refinement to your methods. I presume one of your initial quests was to see how far you could safely extend an OCI; you've done that.

Now, if you want to compare and contrast lubes, you need to have the lube be the ONLY variable, so hold the OCI steady (10k miles, 8k miles, 12k miles - does not really matter as long as you hold it steady). Establish that the lube being used is capable of "normal" results. Once you do that for each, extend the OCI out bit and test for "normal" again. Keep doing that until you notice a shift outside of normal wear variation. You want to slowly extend them out until you get some shift in your "normal" data response. You'll want to run three consecutive OCI/UOAs with each lube. The reason you don't just grab the first one that bounces is because there are events that can throw out a "flyer" and would otherwise skew data. We've all heard of bearings getting a "particle streak", and the Pb or Cu will spike. That is not conclusive data of a shift; that's an anomaly. You need three to show a trend beginning. Once one of them starts out of control limits then you'll know where one is compromised. One non-normal result is not really telling; it's not conclusive. You need to see three to know it's not a fluke. Keep upping the OCI until you note a shift. Once you note a shift, that becomes the baseline for contrast. Both fluids need to be run until they supply a shift outside normal; that way you know the true limit of each lube.

Let me be clear about a few things.
1) when I say "control limits" I am not talking about condemnation limits; those are different. Control limits are what the range establishes; here it is the macro data mean and standard deviation in play. In short, the "normal" variation you'd expect from mass market response. Condemnation limits are what you choose to remove a product from service. Different things.
2) when I say "compromised" I do NOT mean to state that you should (or will) risk your engine. I mean that compromised simply describes being out of normal limits, but not running an engine unto death's doorstep.

Yes - I meant wear metals values in the UOAs.


So to kind of put this in perspective, I'll give a make-believe example ... note these are not the real numbers for your Triton; I'm pulling them out of thin air for the example.
Say your engine family shows Fe averaging 2.0ppm/1k miles at 10k miles duration. The sigma (std dev) might be .3ppm at that exposure. Until a lube would exhibit >2.9ppm/1k @ 10k, it would be considered "normal". And to make sure the data is consistent, you'll want three consecutive samples for each product tested at that one exposure limit. In this example, you would need to see three UOAs in a row, of that one lube choice, at that one OCI duration, exhibit Fe greater than 2.9ppm/1k miles to know the lube had gone outside "normal" capability. Until that happens, you have not compromised the lube performance. So, you keep upping the OCI (the variable) until you find that you can affect the performance output by manipulating the input.

That's what you're after; you're trying to find how far a lube can be pushed and begin to exhibit non-normal response. Then once you establish that for both lubes, you can then apply the fiscal analysis of which is more cost effective. Because you are not doing micro data collection, you cannot apply micro data analysis. You have limited sample data sets, so you have to apply macro data analysis; you can only analyze your data against "normal" mass market data. You have to see how far you can push something past "normal" performance results. Anything inside that normal range is, well, "normal" and the variance is expected.

Hope that clarifies it.

Is it hard, involved, expensive and time consuming? Yup. I never said it would be easy. I only offer this advice to those that are interested in truly knowing how to choose their lube (or any product for that matter) relative to the performance and cost structure.

Everyone is free to have an opinion about what is "best" for them; we'll not deny that. But if you want to really claim something and have me believe it, you have to approach it with the understanding and dedication to properly use the appropriate methodology. Otherwise, you're just stating an opinion and not facts backed up by credible data. This is why I take exception to most folks that look at one or two UOAs, and claim something to be "better" than another. Bad mojo - can't be done. Most often, they only look at the raw numbers; they have zero idea what the running mean is, and even less indication or knowledge of the actual variance of normalcy. Hence, they jump to an unfair conclusion; that is what you and Jim did. I'm not saying that to be mean or spiteful; it's just the truth. You made a conclusion based upon flawed assumptions and bad methodology application.

Anyway - I'm not trying to monopolize your thread. But you and I have conversed privately, and I thought this would be helpful to clear this up not just for you, but others. You are on the cusp of being able to prove some very interesting things, but you need to be careful of the methodology you use, and how you interpret the info. Minimize all input variation; pick one OCI. Establish normal behavior. Increase that one variable (OCI duration) until you detect repeated non-normal response; do this for all tested subjects (lubes). Then apply a fiscal review based upon life cycle to conclude an ROI.

If you recall the info in my "normalcy" article, I mention that there are two methods to compare/contract items. BOTH of them are involved, and need to be done correctly to supply a reasoned credible answer. One takes a long time and a lot of money (macro analysis). The other takes a LONG TIME and a LOT OF MONEY (micro analysis).

I'll bet you're sorry you asked now; I know that I'm long-winded, rude and boring. Sorry.
 
Last edited:
Dave - I am afraid I will have to call "uncle" on that one. My "experiment" was never going to be that detailed and if that means it does not prove anything, then so be it. Hopefully, the data that I have published to date allows someone (it should specifically help the Modular guys) make some informed decisions about oil life and oil choices. There will be a few more OCIs to help me decide on a final oil choice, but I do not think I have the desire to experiment along the lines that you suggest. Sorry!
 
Glenn- You absolutely provided a great roadmap for us 5.4L fans (or fools...it's a love-hate thing), so thanks for that.
 
I think your experiment has great value. I think Dave's point is that if we're trying to compare one approved lube against another approved lube (aside from the easy to compare things like TBN), we have a serious job to do.

Picking a condemnation value and comparing a synthetic to a conventional is fairly easy and straightforward. Heck, you've already demonstrated what "basic" lubes can do. But, if we want to compare protection of one versus another, or really test the limits of a lube, it's a very involved process.

That's why I say we should trust certifications. Look at how many miles you put on and how difficult it would be to prove anything beyond the possibility of extending OCIs. Look at what taxi fleets are able to prove, even M1's taxi test - an approved lube works well. That's not any great revelation.

That's why standardized tests are so helpful to the oil companies and automakers. On the other hand, the data you've provided for OCI extension possibility in your platform is invaluable.
 
The trials you've run so far have great merit. You have in fact exhibited much data in regard to two things:

1) extended OCIs are very safe
They prove that extended OCIs are very doable, regardless of the lube/filter used. They show that contrary to conventional wisdom (I prefer to call it uninformed bias), lubes and filters can go way further that most think and still SAFELY protect the engine.

2) normal results can be found in many products with equal expectation
Additionally, we can conclude that at the exposure limits you've imposed, the lubes all perform the same. It's very much like the transitive property in mathematics. A=B; B=C; therefore A=C. All your results are "normal", so all the lubes performed essentially the same. Until something would push them outside normal, there is no distinguishable difference. In fact, your results are VERY normal, within one standard deviation! Your engine is in fantastic shape and the lubes you've used offer excellent protection. That is not unique to those lubes; many offer such performance. But it is reassuring to know that one's rig is well cared for.

Do not for one second think I don't appreciate and applaud your efforts; I most certainly do. If I gave the impression otherwise then I again offer my regrets, as that is not what I wanted to do. If you choose to stop right here, you've still provided a great service to us all. Thank you for that!
 
Last edited:
I just wanted to say i appreciate the time and effort 2010_fx4 has taken to show us his numbers. While i get where dnewton is coming from,still numbers... are numbers imo. I was on the fence about using afe 0w-20 in my 5.4 3 valve this all but sets the tone for a healthy dose of this mobil 1 flavor going forward. Out with the edge ti im currently running which has been a fine oil in its own right.
 
Status
Not open for further replies.
Back
Top