Unless your refills are all very close to the same # of gallons, directly averaging the mpgs from the MFD is not the most accurate way. One should calculate the gallons comsumed by dividing

(miles driven) / (MFD #) = gallons used

add up all those gallons, and divide the result by total miles driven.

DAS

 

I am still on my first tank on Prius II. I have driven about 400+ miles. The display shows a MPG of about 64 mpg. I still have three bars left in the fuel guage in the display console of the car .. So, I should expect a FE a litttle less than the reported 64 MPG by the car's computer .. huh?

 

Actually, I didn't directly average the MPGs from the MFD, as you'll see if you try it.

Instead, I did ((tank miles * MFD#) + (same thing for each tank)) / Total miles driven.

 

The tire size on the American version of the Prius has a lot to do with the optimistic reading on the display. The tires are slightly larger in the ones soled in Europe and Japan, so they have much less error. I still find that the display feedback helps me keep trying to get better MPG averages every tank.

https://www.greenhybrid.com/compare/.../signature.png

 

Well, I'm up to seven tanks so far, and the numbers are starting to indicate things.

MPG and Avg are based on hand calculations from the odometer & gallons indicated at the pump.
Tgals is Total gallons pumped so far.
MFD is reported straight from the screen.
CGals is based on the number of gallons the MFD seems to think the car used (Miles/MFD).
TCGals is total gallons the MFD thinks the car's used so far.
CAvg is is (Odo/TCGals). Sorry, Tochatihu; you're right, that's the best way.

Odo__Miles_Gallons_Tgals__MPG__Avg_MFD__Cgals_TCga ls__CAvg
_321__321___7.005__7.005__45.82_45.82_46.2__6.948_ __6.948__46.20
_789__468__10.468_17.473__44.71_45.16_45.9_10.196_ _17.144__46.02
1057__268___6.297_23.770__42.56_44.47_43.5__6.161_ _23.305__45.35
1428__371___7.598_31.368__48.83_45.52_45.1__8.226_ _31.531__45.29
1699__271___6.599_37.967__41.07_44.75_45.9__5.904_ _37.435__45.38
2125__426___7.502_45.469__56.78_46.74_45.9__9.281_ _46.716__45.49
2480__355___9.674_55.143__36.70_44.97_46.1__7.701_ _54.417__45.57

Of note:
1. Compare Gallons to Cgals and you'll see the MFD usually thinks I use less gas than I do, but looking at tanks 4 & 6, it seems obvious that I wasn't able to completely fill the tank, especially tank six, since tank seven got REALLY bad numbers. Tank seven, BTW, is where the pump stopped after 3.5 gals, giving me only half a tank, and I had to drive to another pump and add another 8.5 gals. Compare Tgals to TCgals and you'll see it all seems to average out in the end.
2. My lifetime average is just under 45mpg; the MFD says it's 45.57. Not too much of a difference!

I'm just about ready to switch to the MFD's numbers rather than my own. I think I trust it more.

 

This may be unrelated but I have a scangauge and it seems to vary greatly by the type of gas I use. I went from using 10% ethanol gas to 0% ethanol and my scangauge immediately was understating my mpg, calculating that I used 1.5 gallons more per tank than actual. I'm guessing the Prius uses a more accurate method of figuring/measuring fuel volume than the scangauge. I'm assuming the Prius has no way of adjusting the MPG accuracy?

Just curious, has anyone see this type of error with different alcohol content?

 

I've been tracking my 2005 Prius mileage on a spreadsheet for 15 months now. I calculate both ways - MFD and pump. For my car, the MFD is 5% too optimistic in the long run. Occasionally you will see 'adjacent' tankfuls that show radically different mpg based on pump calculation; that is the variable cutoff effect. But as has been explained here, in the long run the pump method must be the more accurate.

We also have a 2003 Prius, and I did the same with it for 1.5 years until I handed it over to my wife, who didn't want to continue the experiment. Anyhow, over that time, it too was 5% optimistic.

While we're on the subject, I've run regressions on this data set that shows that mpg increases with ambient air temperature, but less so as you approach higher temperatures, so it basically levels out at 95 F. But the difference between 10 F and 30 F is pretty big! Also, the gallons that will fit in the tank until auto-shutoff increase with temperature as well. I almost always run until the last bar is blinking, so my tankfuls tend to run 8 to 10.5 gal.

 

My 2 cents...


I also post MFD numbers in the comments section of the DB like FooMonkey does.

MFD is ~4% optomistic over 25+ tanks with the same tank to tank extremes mentioned earlier.
I agree that there are probably multiple things ocurring that cause this like ECU sampling errors, temp caused bladder fill rate changes, pump discrepencies, and auto shut-off differences. Let's see, that's for items contributing a total of 4% - not bad.