MFD MPG Displayed Number - Could this be fairly accurate?

I've also tracked every tank and just yesterday updated the data in my spreadsheet. The car now has 10k miles on it and the numbers are:

Actual average MPG: 44.9
MFD average MPG: 46.0

So in my case the MFD is 2.45% optomistic. Seems in the ballpark to the findings you're getting. I get my numbers a little differently than you, but both ways should be equivalent. What I do is use the MFD MPG per tank to calculate the gallons the car think it used. Then to get the average MFD MPG I add up all the gallons the car think it used and divide that into the actual miles driven.

 

I've typically shown about 2 mpg higher on the mfd than what I get in the calculations at the pump over time.

I don't feel this is error in having filled a little more or less on any particular tank since I see this 2mpg difference pretty much every time. Sometimes a little more than 2mpg error, sometimes a little less.

The last 6 or 8 tanks have typically been actual mpg 52 - 54. The mfd has shown 54 - 57 when I pulled up to the pump with about 500 to 530 miles on the tank.

http://d.spicher.home.comcast.net/prius/mileagegrid.html

 

The vehicle computes average FE using a fuel metering method over distance traveled, or rather, fuel through injectors vs distance. This is a superior way of accurately determining average FE vs. the fill up method that we employ when we are getting fuel.

There is a great deal of variability in fill ups, especially with the additional factor of the bladder based fuel tank. I'm convinced that variability in fill ups makes the MFD look optimistic, however I don't see any reason to doubt the MFD considering that the onboard computers have full knowledge of distance and fuel metered. I on the other hand am using gross numbers when calculating FE by hand after a fill up.

 

Also, don't forget that temperature variance can affect bladder volume.

If indeed the Prius calculates MPG from measured consumed fuel, there is NO DOUBT that the MPG calculations are dead on accurate and much more accurate than hand calcs. But, I keep a running log of both measurements everytime I fill up.

 

Since 2001, the Idaho National Laboratory's Advanced Vehicle Testing labs have put 28 hybrids through 1.6 million miles of real-life road fleet testing, as well as an extensive battery of lab tests. It's part of the Department of Energy's FreedomCAR and Vehicle Technologies Program.

They state, “Francfort and his team also determined that the computer dashboard readings for fuel economy—which most drivers use to base their judgment rather than manually calculating the actual mpg—inflates mileage numbers by somewhere between 10 to 20 percent.â€

Looks like the Prius is better than most.

 

Does anyone know how often the comp records a data point for MPG calculation?

 

I was disappointed a few times when my measured mpg was about 10% lower than the computer calculated, but at my last top-off the computer was saying just below 50 mpg and the bladder didn't accept much gas so my hand calculated mpg was over 60. (163 miles over 2.7 gallons; the bladder would have to accept about 3.3 gallons to have agreement with the computer)

With my short commute and limited weekend driving, I'm wondering if I should take it in to change the gas from time to time.

 

<div class='quotetop'>QUOTE(San_Carlos_Jeff @ May 16 2006, 10:07 AM) [snapback]256288[/snapback]</div>

How do you know that the MFD is being optimistic? You're calculating your MPG based on how much you put in the tank. That doesn't necessarily have any relationship to the amount of fuel you actually burned. How much you put in the tank is a function of several variables so there's a lot to account for there. I wouldn't label it actual avg mpg. I'd label it manual avg mpg. There's no way that anyones manual calcs reflect reality everytime they fill up, especially in a Prius with its testy fuel bladder.

 

<div class='quotetop'>QUOTE(tripp @ May 26 2006, 11:01 AM) [snapback]261330[/snapback]</div>

You're correct, tank to tank manual calcs are pretty worthless with this car. I don't even bother doing that anymore. However my manual calcs are an average over 10000 miles (actual cumulative gallons used divided by total miles driven). The only way the bladder could affect this number is on the last fill-up, something that will have little effect spread out over more than 10k miles.

The only way my manual calcs can be inaccurate is if the readings on fuel pumps are consistantly wrong. Since fuel pumps need to be regularly calibrated I'm hoping that doesn't happen.

 

<div class='quotetop'>QUOTE(San_Carlos_Jeff @ May 26 2006, 10:12 AM) [snapback]261336[/snapback]</div>

Season temp flux could also have a significant impact over the course of the year. Fuel pumps are calibrated but there seems to be a significant difference in the sensitivity of them to premature shut off. Unless you always use the same pump (and maybe you do) you can't get any consistency in your dispensation. That alone would seem to cast doubt on the manual calc methodology, regardless of how many times you fill up. It's just not a repeatable experiment, whereas the MFD is.

 

FYI:
I'm only on my third tank, and I 'reset' the consumption screen with each tank. My first tank was 366 mile, 43.2 MPG vis 44.2 on the MFD. My second tank was 48 MPG vs 50.4 on the MFD. Reasonably warm spring weather, driving mostly short trips with some curves and not completely flat, but not 'hilly' by any means.

 

Over time, how much gas you put in the tank (including the bladder effect) makes no difference. You could just fill half a tank if you so choose. So long as the pump and odometer are accurate, any discrepency in one tank is accounted for in another. As you have a larger number of data points (tanks), you reduce the statistical deviation and at some point it becomes insignificant. This makes overall MPG calcs much more accurate than other methods.

 

<div class='quotetop'>QUOTE(tripp @ May 26 2006, 11:32 AM) [snapback]261346[/snapback]</div>

None of that matters when figuring out the avg. MPG over many miles. For the first 10,063 miles on this car I used 224.4 gallons of gas, to give me an average MPG of 44.9. All the other variables don't matter.

 

gordonr,

My error rate is very close to yours. And as mentioned, anyone who either panics or jumps for joy after "measuring" MPG based on any given single tank doesn't know the Prius's foibles.

In fact, those of us who owned '01-'03 Priuses (sometimes called "Gen 1") know that the error rate was worse, and that Toyota corrected the previous MPG tracking rate starting with the '04s. We used to see discrepancies of 4-5 MPG all too often.

 

<div class='quotetop'>QUOTE(gordonr @ May 16 2006, 08:50 AM) [snapback]256275[/snapback]</div>

Check out John's web site. He keeps extensive records and might have the comparison you're looking for.

 

After 6,000 miles I'm averaging 49.84 and the MFD is averaging 50.14.

 

Mine is a bit pessimistic, but it is really off if the temp goes up, down, up again. I don't know why but I get appear to get greater variance when we get big temp shifts is the spring and fall.

 

<div class='quotetop'>QUOTE(hdrygas @ May 27 2006, 05:55 PM) [snapback]261888[/snapback]</div>

Shifts in bladder capacity between fills?