Discrepancy in Average and Calculated MPG

Not bad at all. I wish the Gen3s were that close. Ours typically have 2-3X larger error.

 

There is only ONE method to determine MPG and that is Miles driven divided by gallons used; aka: Pencil and Paper, just as you have done.

The ECM uses miles driven divided by CALCULATED fuel used based on theoretical fuel flow (liters/hr * hours of injector PW), which inherently introduces tons of error.

 

Well, speedometers/ odometers have a variance to actual.

2.5% variance seems in line with a mileage to gallons used.

As for your question, IMO probably both mileage calculations are presented by forum co-members using the statistic they find easiest to report on. For example '16 V average screen shows about (about) 41 mpg. I do not know the comparison calculated mpg.

Great fun figuring fuel mileage!

 

We watch a lot of detective shows in the aft, and "motive" is a big factor in solving cases. Now who would benefit from exaggerated display? Toyota?

I agree with @fuzzy1 : our third gen has an unbudgeable lifetime error around 7.5%. It jumps around a bit, due to pumping vagaries, but 7.5% is where it sticks.

 

perry - I find that the "on-screen" mpg is usually a little optimistic but the variance actually works both ways. To give you a comparison, here are the results from my last 10 fills:

10/09/2016 - on screen 41.50, calculated 44.07
09/26/2016 - on screen 46.20, calculated 40.76
09/18/2016 - on screen 40.80, calculated 40.36
09/10/2016 - on screen 42.10, calculated 44.94
08/29/2016 - on screen 37.30, calculated 34.34
08/14/2016 - on screen 44.90, calculated 46.73
08/13/2016 - on screen 42.70, calculated 48.48
08/01/2016 - on screen 43.10, calculated 37.39
07/31/2016 - on screen 45.10, calculated 46.48
07/27/2016 - on screen 44.10, calculated 42.02

We don't have a regular fuel tank and the amount the tank will hold changes based on the flexibility of the tank's rubber bladder at any given time. I don't depend on the displayed mileage and instead just fill it up when it's convenient anytime after I see two bars left on the gas gauge.

 

Man, that jumps around a lot, probably due to the bladder tank, as you say.



(bottom right number the average diff)

 

You can erase that tank variability uncertainty by keeping track over many consecutive tanks, and computing the longer term average.

 

^ This.

Vagaries abound on single tanks, but the more the tanks you accumulate, the more accurate (and unbudgeable) the results. In other words: calculate individual tanks, say in a spreadsheet. Then off on the side do a calc that uses the total gas and total gas (typically two columns). That number will reflect the cumulative mpg, regardless of the ups and down due to varying tank capacity.

That's why it'll take another year of sub 4.7 (liters per 100 km's) to knock my 4.9 Fuelly badge down to 4.8, lol.

 

I would a spreadsheet too, if possible.

 

Most "modern" vehicles have ODOmeters that are MORE accurate than GPS, speedometers not quite so much for reasons I do not understand - it should be simple.

All modern vehicles are programmed to "pulses per mile", this is simply PPM = (# driveshaft teeth) x (axle ratio) x (tire revs/mile). When you start multiplying this out, you get big numbers like PPM = 16 x 4.34 x 613 =~42567.

Let's say you "gain" 3 Revolutions Per Mile on worn tires compared to new. 3/613 sounds meaningful, but it's really ~0.5%, so you can't really attribute MPG changes to differences in rolling diameter of the tires. Inflation from the lower to upper threshold of a tire can affect rolling distance on the order of 2% or more and in fact, Ford used that to determine the potential for an underinflated tire for a number of years before TPMS sensors were affordable.

Because the number of teeth and gear ratios are EXACT measurements that are elementary to make, the only variable is rolling tire circumference (aka: revolutions per mile).

Because SPEEDometers also factor in time, a measurement with *some* error (more or less depending on the quality of the clock), the accuracy of the speedometer is inherently less accurate.

If you want to verify your odometer, find a 10 mile stretch of highway with mile markers, drive it both directions after resetting the TRIP meter as you cross the starting point. If accurate, the trip meter should register "exactly" 10 miles; a one block error is 1% and typical. To check your speedometer, run the same course at 60 MPH indicated, you should cover the measured 10 miles in "exactly" 10 minutes. As always, convert to decimal and do the math for your actual speed in MPH.

You don't KNOW you MPG unless you know your MILES and with more miles and gallons the more accuracy.

 

Time can be measured more accurately than anything else, and 0.005% accuracy costs just pennies. For these purposes, time measurement errors are negligible compared to other errors.

Part of the speedometer display 'error' is intentional, to comply with laws and industry standards forbidding it to read low. Because other measures in the calculation have some uncertainty, an error distribution is present. To keep units on the low side of the error distribution in compliance with the requirements, the system is biased high. Euro-market vehicles are commonly biased higher than U.S. market vehicles, originally by industry regulations, now mostly just to keep in line with their respective histories.

Odometers are now commonly biased to read low. If they had the same high bias as the speedometers, product liability lawyers would swoop back in with more class action lawsuits claiming warranty fraud. The newest car in my household is biased noticeably low, after its predecessor was included in such a suit and settlement. (My particular car did not display the alleged issue.)

 

Editorial: first word in last paragraph was meant to be Odometer?

 

Yup. Fixed.

Low blood sugar today as I prep for a common senior 'test' tomorrow.
That's my story, and I'm sticking to it.

 

Agreed, but they don't, two small errors multiplied, make a larger one and that is why SPEEDometers have greater error in most cases than ODOmeters (they can have errors in opposite directions).

Not. There are no laws that that forbid an Odometer nor Speedometer to read low as you assert, OEMs have for YEARS worked to get greater and greater accuracy. In point of fact, both on our 2016 Fusion read low, very slightly so(<0.5 block in 10 miles), but low, based on multiple physical measurements. I will agree that it is more common to a speedometer to read high than low, but over millions upon millions of miles, any trucking industry professional will tell you that odometers tend to read low.

 

Mine has always been almost exactly 3mpg under whatever the readout says.

 

Speedometers typically read 1mph higher than actual speed so that you are not speeding without being aware. Meaning if you drive 60 mph for one hour, you may have only traveled 59 miles, making the odometer reading lower than the speed displayed on the speedometer. Hence the inverse correlation.

 

We traveled southern California-New Mexico-Colorado-southern California this summer as well as southern California-Salt Lake City, UT-southern California a couple of weeks ago. We don't care for the Toyota GPS so we had a Garmin GPS unit with us for both trips. We found there was less than 1/4 mph discrepancy at multiple speeds when we compared the speedometer to the Garmin, with the car being 1/4 mph higher than the Garmin.