Computer vs. computed MPG

In this thread, I posted my results over the year 2005. I found that the mileage reported on the MFD was 2.73% higher than than computed from my written fuel log. This amounted to about 8 galllons of gas over about 15000 miles.

I did come up with several possible explanations for the discrepency not all of which point to chicanery on the part of Toyoto:

1) Evaporation from the tank--not likely due to the bladder

2) Inaccuracy in measuring fuel flow from the tank to ICE, leading to a error on the MFD side

3) Commerical fuel pump not accurately recording fuel added. This would be related to how fuel was added. After auto-fill had stopped, I ususall managed to get 1.5 to 3 gallons in in small portions. I never overfilled the tank, but it is quite possible that despite certification, the gague at pumps may read high for these short hits. This year, I am deliberately not trying to max my fill, but actually shut down soon after the auto-fill stops (the top bar on the fuel gague now disappears in about 80 miles, as opposed to 120 miles before). I'll see if this makes any difference when I again do a summary.

 

There is another possibility which probably would not account for all of the difference. That is that the pump is not delivering exactly as much as it says it is delivering. There is an allowable tolerance for inaccuracy. Very few pumps are absolutely accurate. Some are very innacurrate and are the cause af many headaches. This is usually a problem only if one pumps from the same gas pump all the time and it is one of these badly inaccurate pumps.

 

And then there are the dishonest station operators who are good buddies with the weights and measures inspector... I've got one in my town. According to their pump, I put over 17 gallons in a 16 gallon tank. There is no initial on the cert sticker, except when I complained and a "surprise" (yeah, right) inspection. The initials disappered a week later. I trust Toyota more than my local stations!

 

gee. you trying to tell me my MFD is wrong despite it varying (and LOWER than my calculated figure) less than a fraction of a mpg??

that is disappointing to me. i just thought i had done a good job of using pumps that were accurate.

 

Mine has generally been a bit higher than the computer, but over time they are converging that is after a year and 3/4.

 

for what its worth, I have meticulously tracked every gas purchase since I puchased my Prius, and have kept track of every useful piece of info in an Excel spreadsheet.

Hear is what I have found:
Miles driven - 29,230
Gallons gas purchased - 572.68
Actual lifetime mpg - 51.04

WEIGHTED average of all MFD readings - 51.40

Virtually identical. I would ask, are you sure you are comparing actual lifetime mpg to a WEIGHTED average of all MFD readings? If you don't weight the mfd readings based on the # miles driven, the comparison is meaningless.

 

i calculate LT mileage by determining the amount of gas used for each tank both calculated and based on the pump reading.
iow, if i went 500 miles and the MFD says 50 mpg, then i calculate what the MFD says i should have used, using tank mileage (for this i use trip A to get a more accurate reading in tenth mile increments. both trip A and MFD are always reset manually by me at the same time always when the car is at rest) divided by mpg's to determine calculated gallons of gas used.

using the method based on the MFD, my car has burned 423.798 gallons of gas. according to the gas pump, i have put in 423.173 gallons of gas. if you read my signature, you will know that the above difference equates to 1/16th of a mile per gallon or .06 mpg * 423.173 gallons or about 25 miles.

ya... i guess if you look at it that way... 25 miles is pretty inaccurate after all. i am bummed now

 

I'm curious if you're in the rounding error range. So when your MFD says 51.4 mpg, I take that to mean from 51.350mpg to 51.44999999...mpg. And the same thing for miles. So, if you add .049 to each mpg calculation and subtract .05 for each mile reading (although the miles wouldn't get reset and would seem to mostly even out over time), how's it look? What about the other way around?

 

Both my wife and I say as long as we get 50 MPG or more we're happy campers! I find that my manual method is higher than the computer more than not. I have NO complaints. Saving at least $20.00/wk. on gas. That's almost what my interest on the loan is this month. (I hate to pay a bank any interest!).

 

In my last refill I found that there was a difference between my milage on the MFD and my trip odometer. While the MFD was showing 412 the odometer was 412.7.

Naturally this caused a discrepency when I did my calculations.

Any yes, I have been tracking every fill up and am using a simple excel
spreadsheet to track gas purchases..........wife thinks I'm nuts.

 

There has been no input into this thread yet from someone outside the USA, where the gallon is only 5/6th of the gallon elsewhere. It would be interesting to have a comment from someone (such as in Canada) where calculations are in L/100kms. I haven't had my Prius long enough to make a valid comparison yet but I do wonder sometimes whether my Prius knows that a British gallon is considerably more than a US gallon.

 

Canada, eh

 

How *a-boat* that!

Perhaps I am committing a sacrilidge but it occurs to me that mileage calculations are best focussed on comparisons to previous readings from the same car using similar methods to what was used in the past what ever they were. There are so many external variables including the types of measurements, the calibration of the pump, the temperature of the gas in the pump tanks, the care taken in pumping the gas, the care taken in the calculation, the mechanical condition of the car, the way the car is driven and on.......

But its fun when we know that we have such a superior car!

Allan de

 

I have a hard time believeing that the Prius can accurately gage the fuel that is being burned as the car is operated, but just to add my anal-retentive numbers to the mix....

46,524.6 cumulative miles
941.973 cumulative gallons
49.39058763 cumulative miles per gallon

50.3 average computed mpg

 

Hi All,

With regard to the car's computation of mileage - measuring the rate of gas flow can be done accurately. Not sure if the Prius is doing it accurately, but technology wise its very doable. One device would be a constant volume rotating impeller wheel with a magnet. The revolutions detected by a hall sensor. In mass production this is a couple pieces of plastic, metal axle and a low-cost magnet and hall sensor, and cheap. There are also ultrasonic flow measuring devices with good accuracy.

The one thing the Prius probably is not doing with good accuracy is the speed of travel. This is due to varying tire pressures with temperature. The efffective rolling diameter of the radial tire is going to vary with tire air temperature. Tire air temperature varies with both enviormental conditions and rolling history since starting at a condition where the air in the tire was the same temperature as the pavement. This assumes the tire does not leak. But it does slowly. Another variable. I believe the Prius only measures average revolutions of the transmission output (between the left and right drive tires) as most cars do.

Rubber in general has variable flex modulus (how much force is required to flex the rubber) with temperature. So, tire temperature greatly effects mileage besides the measurement of mileage. Running with Tire Pressure above the recommendation woud also effect the actual mileage and the measured mileage. Higher pressure should improve the actual mileage by reducing the amount of flex. If the measurement is by tire rotation the larger diameter of the overpressure tire would make the mileage read worse than actual ( gas/rotationlength - actual rotation length now longer than assumedrotationlenth used by computer).

A ground directed doepler speed measuring device (Ultrasonic - not dog friendly, IR laser radar - lidar - not desert friendly, Radar ?) might be more accurate than counting tire revolutions. I have not seen any reference that the Prius has such a thing, however.

The 2006 Prius has tire pressure monitoring. One could map tire running diameter with the pressure, but this assumes a specific tire. Is there any setup variable in the diagnostics for this ?

 

The jury is still out on how much of a speed delta you actually
get on a soft tire. Differential-speed systems seem to be taking
a back seat to RF-based direct pressure/temperature monitoring
systems built into valve stem bases or sensors strapped around
the rim. At some point this summer I'll be experimenting with
the differential concept to try and finally resolve this, and even
that might only be meaningful for certain types of tires, sidewall
stiffness, etc. And if they're *all* soft, you lose.
.
_H*

 

I don't understand why people bother with the manually logging and calculating the milage. Lots and lots of people here have stated that the MFD average is higher than their calculated average, however they all state that it is high by a consistent amount and less than 2 or 3 MPG.

So back to the question of why people bother to manually calculate their average... I can understand the desire to monitor you gasoline usage and your efficiency. But isn't it the trend and not the absolute value that is important? Is it really important to know that you really got 51.5 MPG instead of 52.7? It seems to me that what you really want to know is "am I getting more or less MPG than I got last month?". Since I believe that it has been concluded that the MFD calculated value is consistent, even if it is a couple of MPG high, then it would seem to me that manually logging and calculating your MPG is unnecessary. It also seems like the MFD should be more consistent, tank to tank, than manual logging, since you can never be sure just how full you filled the tank, and even a quarter of a gallon error can make a big difference on a tank to tank basis.

But then again, I've always been a bit lazy and that kind of meticulous tracking has never appealed to me. Maybe some people just enjoy that kind of charting activity and if it wasn't MPG, then they would be charting the daily air temperatures and the growth rate of their grass in mm/day.

 

You clearly don't understand the anal retentive mind!!
Certainly it is true that you can not compare a single miles/gal fill up with the computer MPG and say FACT! Only with longer term logging will you know the true difference between the weighted computer calc and actual miles/gallons. My computer calc is about 2% generous. Now, the only REAL value to this is being able to use the computer's calc at the pump to figure out how much you should be able to get in the tank. You can always just stop when the pump clicks off, but the back pressure of the bladder seems to shut it down early (about 8.5 gal even when the calc says I should be putting in 9 or more). I don't like to stop for gas any more often than necessary and I don't want to spill gas which could happen if I just trusted the raw number of the computer calc.

 

This an intriguing topic/thread. Depending on what we are looking for we decide how to seek it. My point in measuring is related to an effort to consume less gasoline. Therefore I measure calculated MPG by dividing gasoline bought by odometer miles per purchase and lifetime. The lowest MPG (40) I have gotten is double what I got on my last car (20). The highest I have gotten is 50 or 2.5 times what I was getting prior to Prius(PP). The average MPG is 45 and the mode is 44. My goal was to buy less gasoline not to argue about how to measure Prius MPG. That as I suggested at the onset of this message depends on what you are seeking.
h34r: