Minimum amount of miles for a good analysis?

[Patman]

I was just curious as to what is the minimum amount of miles needed in order to get a fairly accurate analysis of how an oil is going to hold up? Is 2000 miles enough to see if an oil is doing it's job, or is it better to go closer to 3000 miles? I have seen a few oil analysis reports on the net where people have only driven 1000 miles or less, and wonder if they are really going to get an accurate representation of how that oil is going to hold up in the long run.

 

[Terry]

Depends on what we are looking for and what type of engine/equipment.

For a passenger car 3000 to 5000 miles is a good starting point.

For wear comparisons 2000 to 3000 miles are good for keeping your test within a reasonable time period.

Key is comparing apples to apples, very difficult in real world as variables change and most do not keep good enough records.

 

[widman]

I agree that you are not really going to see much about the oil until you have put 3000 to 5000 miles on it. And then only after the 3rd oil change. The first 2 oil changes still show trending downwards of the stuff left behind. One of these days maybe I'll sit and consolidate some of that, but I've seen it many times.
All this being said, I sampled a "famous brand" for a customer last month at only 800 miles, assuming that it would look good. It had more wear metals, less dirt, and less additives than the other brand at 6,000. It had also dropped out of the 15w40 range.
Bottom line is if the oil is bad enough, you can see something early. Otherwise, get some miles on it, and 3 oil changes.

 

[Patman]

I guess part of my reason for asking is that in my neverending quest for knowledge, I want to try out different oils and do analysis on them. I do realize that the best way to do it is to establish a trend with a few oil changes in a row, but that's not entirely possible, it'll take much too long that way. This way will at least give a general idea of which oils are doing a better job in certain areas.

 

[ToyotaNSaturn]

A worthy topic, time to resurrect it.

 

[HondaMan]

TNS,

I think this is a very worthy topic.

I have been taking UOAs @ 3.5k OCIs, which most see as a total bore, so that I can establish, for myself, what I believe to be a nice baseline and trend.
I feel the interval is just right as mentioned above.

I have a UOA kit in hand which will be a total of three with the same oil, filter and OCI mileage.

This next fill will be different for me how?
I intend to use LC and NOT take it out a million miles, but take it out to the same 3.5k

I feel this will actually be a nice contribution to the board in that it does take TIME as Patman mentioned almost four years ago and a CONSISTENT approach. And after I take it out 3.5k, we will see what changes/improvements where made.

Again, just takes awhile as Patman said.

 

[TomJones76]

Wow. That is a real blast from the past.
I can't IMAGINE Patman asking that question today. He went from being a student to an expert, albeit not on Terry or Molakule's level.
And the best time to sample is every 1000 miles over your entire interval. Of course you'll want to drain oil off the topside with a pump, so you don't drain TOO much and skew your results.
I would *kill* to have the money to do that, so I could make nice little charts of my oil's progressive thickening and unthickening.
If you CAN'T do that, I think the best time to sample is at the highest interval you're comfortable doing. I see guys drain and sample early but it is such a waste to drain Mobil1 at 3,500 and sample it like some of these guys do. Unless your engine is blowing up, that M1 will be PERFECT at 3,500 miles.
Draining at 1K is too early. Oils do this *thing* between 0 and 1000 miles, and the initial "stuff" loading into the oil is not a noteworthy process. You start to see the oil's real trends developing around 2000 miles.
Sampling at 2,000-4,000 miles gives you an ideal of the performance of the oil in light usage and short drains.
If you want to see how much endurance it will have, 7500 to 9000 miles with TBN is a good test.
If you're trying to see it fail, run it for whatever 12 months is for you, or 12-15000 miles and then sample with TBN. Now THAT's a test right there.