Man, Buster's really got it out for me with this top-up thing.
I dunno what to tell you that will placate you bud, you've seemed to have made up your mind on this one.
My responses are as follows (to be added, retracted, amended, or disowned as I may later find convenient):
1. the primary purpose of the test is to determine the *relative* benefits of one motor oil vs. another. As long as all oils undergo the *exact* same test regimen, this primary purpose will still be valid.
1a. any other thing we learn from this test is bonus.
2. The 1/12th of capacity that I periodically have to add ages along with everything else. So at 6k miles, the 1/2 quart added at 3k miles is 3k old. Therefore to just add up all makeup oil from 3k to 6k and say "well crap it's got 1/3rd fresh oil in it!" is erroneous. Indeed I would suggest it ages faster than the other 5.5 quarts as it is quickly diluted and the elevated contaminants in the used oil will break it down faster.
3. The adjustment to TBN for makeup oil is easily calculated, based on work documented in an SAE paper that I have. At the end of the test I will provide calculated TBN values for both removing the samples and removing all oil consumption completely.
4. Particulates are expressed as a ratio so re-calculating the values to compensate for samples is cake. Expect to see some analysis of this at the conclusion of the test.
5. Viscosity has only shifted by 5% even after 13k miles. I find it doubtful that make-up oil alone is causing this. Note that viscosity did NOT improve even after adding 1.5 quarts at the filter change.
Then there are some other bonus material we've learned:
* we've seen for ourselves how TBN drops sharply and then flattens out.
* we've seen for ourselves that OTC oil filters work very well.
* we've seen for ourselves how risky it is to base a decision on single-point oil analysis, due to the fluctuation of numbers.
* and that winter driving doesn't always severely increase wear numbers.
To say nothing of all the interesting tangents we've pursued along the way.
Of course it's not a perfect study. None are, especially ambitious ones. But we can lay out the assumptions we have to make up front, and figure out what values we can deduce, and be honest about the limitations, while still finding something of use to us in the results.
Let's face it, part of the reason this study undergoes such scrutiny is that every data point and every process is laid out for public view. In Mobil 1's durability studies, do they give you every tidbit of information about how the test is performed? No they do not. They say "hey we made it a million miles!" And you have to take it on faith that they didn't fit the engine with, you know, a 20-gallon oil pan or something.
(a little hyperbole to lighten the mood there, I know they probably didn't really fit it with a 20-gallon oil pan). I realize this is beside the point, and I don't mind being put under the microscope, but I just wanted to draw your attention to the fact that other tests may face similar limitations, you just don't know about 'em.
I like mph's table but it needs to be recalculated to reflect actual oil consumption. With the right numbers in there I may include something similar in the final analysis.
Wow, it's getting late, and I'm rambling on. I won't be available Friday but I'll be sure to check in on Saturday to see how our heat:light ratio is doing.
Cheers, 3MP