MPG vs Miles on Oil

[surfstar]

Some people have anecdotally stated that "my MPG decreases as my oil gets more miles on it", and I would prefer to have numbers to put with the claims. Luckily I got numbers, tons and tons of numbers. I went back over my MPG data from 120k – 180k miles, as the numbers have the least variables, including the same set of tires for that time. I then looked at my oil change dates/mileage and put them together and looked at the MPG related to how many miles were on the oil. The first set was interesting:

Miles on Oil / Average MPG / count
0000-1500 / 33.93 / 42
1500-3000 / 33.81 / 48
3000-5000 / 33.65 / 60

The MPG does seem to decrease as the miles on the oil increase. So I broke it down farther and got this:

Miles on Oil / Average MPG / count
0000-1000 / 33.93 / 27
1000-2000 / 33.91 / 30
2000-3000 / 33.76 / 33
3000-4000 / 33.67 / 29
4000-5000 / 33.74 / 29
5000-6000 / 32.11 / 2

Unfortunately you can’t give credit to the 32.11MPG average for 5-6k miles, as there are only 2 data points, which is not nearly enough to allow for variance. It almost appears as though MPG DOES decrease with more miles on the oil. Although the variance between them is so small that it is not significant statistically.

Other points:
Oil was Mobil 1 10w30 for all but two OCI, which was a mix of Mobil 1 10w30 and EP 5w30.
Yes I do like playing with data sets.
And yes I do have a lot of spare time at work. Here’s to BITOG

 

[Al]

Well good work on your part. One could argue that the data is not controlled in that conditions vary from beginning to end and its all about statistics and confidence levels...its interesting, none-the-less.

 

[Ugly3]

Great work! I would love to see this data on an oil that tends to thicken over time, bet the MPG would be way down. The 3MP synthetic life study indicated a significant drop (maybe 10%) as the oil ran into and past the 12,000 mile range.

Thanks for the info!

 

[lght1]

Hi

I keep mpg records each day on a spread sheet. Oil changes and any other maintenance items are mentioned at the day and mileage they occurred.

It is possible that i could extract the data for many thousands of miles over weeks and months.

 

[Shannow]

surfstar, my 4Runner used to have a distinct turndown at 12,000km.

Here's the trends of my Navara (3litre turbodiesel)

 

[kanling]

Surfstar, since your MPG numbers monotonically decrease with age... I tend to think that the trend is real and not in the statistical noise.

0.2-0.3 MPG difference between old and new oil though. Not enough savings to cover more frequent oil changes.

Good analysis and good information.

I wonder if the reverse would be true with conventional oil, which would tend to thin with age rather than thicken?

 

[surfstar]

The data was for 150 tanks of gas, so thats pretty extensive and although there is the .2-.3 decreasing trend, kanling you have to look at the standard deviations of these averages, which are all over 1.0. So although the averages look like they might decline, the raw MPG data ranges from 31-36.

My OCI with M1 is only 5k miles, and I'm sure that longer OCIs would show a trend.

Shannow, in your graph above I don't see anything that would indicate a real trend. The fluctuation of your MPG is within reason for tank to tank variances.

It very hard to find real differences in real-world MPG data, although I know for sure that the combination of my new tires and winter fuel starting on 11/1 this year have caused a consistent drop of 3-4 MPG (over 6 tanks each one showing the decrease), which is why the MPG data with my new tires was excluded.

 

[Shannow]

sufstar,
precisely, on this car I've seen no trends other than different averages on different oils.

 

[peterr]

I am not arguing that this could be happening and think it is anecdotally interestging but I would tend to think that the variations are so small, and the variables too large, and the sensitivity of measurements too loose to draw firm conclusions.

I mean even something like "I am more likely to change my oil when its warm out" could skew numbers by more than 1% just based on temperature and gasoline formulation.

again, not trying to be a naysayer, and its entirely reasonable that you have documented real effects, I just don't think that in a statistical sense you can say this is proof.

 

[oilyriser]

My car definitely idled more smoothly with fresh vs 7500 mile old oil. The old oil had thinned to 9.2 cSt, yet when I had used Redline 5w20 of the same viscosity the rough idling did not occur.