mileage discrepency

I have put 10000 miles on my 2005 and the screen and actual are off more often that not perhaps because we are not filling to the same point. After getting doused with gasoline a couple of times I gave up trying to top off the gas.

 

As I recall from many threads a couple of years ago when the fuel gauge vs. fuel tank discrepancy was really a problem (sorry, too lazy to look them up), the generally accepted agreement between indicated and calculated MPG averages was about +3-5%. In other words, the average on your MFD is probably 1 to 3 mpg more optimistic than you would calculate from the gas pump and mileage between fill-ups.
Having said that however, on a 3700 mile road trip this summer in my 2004 Prius, when I actually bothered to keep track of the distance and fill-up amounts (I had learned from my 2003 Prius that calculations based on the pumped amount were so unreliable as to be useless), the discrepancy on any 1 of the 10 fill-ups varied a lot: from +16.9% to -7.2%. The overall discrepancy averaged over the 10 fill-ups ended up at +4.4% (or 51.7 indicated vs. 49.4 calculated).
A couple of the most important variables are the temperature of the bladder inside the tank and the atmospheric pressure. These affect how flexible the bladder is and how willing it is to expand to "full" (which is a constantly moving target). Combine that with variations at the pump which include how sensitive the backpressure cutoff sensor is and how honest the metered amount is and you get an idea of why the variability is so great.
Anyway 5% is better than average for onboard fuel calculations in most cars, at least in my experience.
Hope this helps.
Doug Wilkinson

 

The onboard computer is retarded. It doesn't know how to accurately measure the amount of fuel in the tank.

 

Some have stated that Toyota has manipulated the computer, so that the Prius appears to have the MPG as rated rather than actual MPG. As far as the bladder having an effect, I believe it does not! There is a fuel sensor that relays the flow of fuel to the computer. The computer then formulates the results to the MFD.

 

The system measures the amount of fuel injected into cylinder 1 and multiplies this by four. Then it divides miles travelled by this figure to determine mpgs. The system is supposed to be quite accurate. Why calculate mpgs when the car does it for you escapes me.

 

Because, by definition, 50% of the population are idiots! That is to say, 50% of the population has an IQ below 100.

 

Why not? It helps determine whether or not the MFD is actually accurate. I keept rack in my pocket pc - it's easy and it's great for tracking my expenses.

In response to the original post, if you calculate each fill up, I think you'll find that after a number of fill ups, the numbers will average out and be very close.

 

Use the MFD as a guide to help you move to better MPG. The per tank mpg both by the MFG and the calculated method are not totally accurate on a tank by tank basis. Keep your own record! The life time average is the best measure of how you are doing. You could do a seasonal average to help out but one tank is problematical.

 

[deleted as pointless]

 

this statement is so wrong. i wrote a blog about this and this opinion is simply a lack of experience with the car.

a new Prius will have a bladder that is new and will slowly expand as you refill it just as anything new will stretch and expand as its being used.

sooooo. you drive around MFD says 50 mpg...500 miles. during that time the bladder stretches to where it is now a gallon bigger inside...no wait... too much, too unrealistic. lets make it a tenth of a gallon.

so it now take 10.1 gallons to refill. on paper that makes the "paper and pencil" average 49.5 mpg...now although the difference is only a tenth of a gallon, the difference on the mileage is a ½ mpg. if you have tracked your average the whole time you have had your vehicle, i think you will find it rare that the difference will be this high for a long period of time.

so we have two measurement scenarios that are liken to the following

scenario one or the paper and pen method...

an accident happens and one witness gives a statement to the police. now the guy giving the statement is smart, alert at the time and had a good view of the situation. now not the best comparison for this purpose but good enough.

scenario two...MFD


a camera captures the accident on tape.

so on the one hand, we refill the tank after the fact, use proven methods of math, make the assumption the pump is accurate, then figure out how much we MIGHT have used since this all we can do.

or we have a computerized system monitoring the car and its fuel usage and distance traveled EVERY SECOND the car is on.

i thinks its obvious what is the most accurate here

 

read my post above

 

And you're signature also shows that the two even out over time.

 

FWIW, I have had my Prius since April 2004. The long-term calculated MPG is 48.5, and the long-term computer indication is 49.5.

However, at each fill-up, I notice that the differnce is larger, and on an indiviual tank, I could see a discrpancy of several MPG. Sometimes the computer display shows a larger MPG, and sometimes my calculation shows a higer MPG.

One way of expressing this is to state the the computer display (in my car, anyway) has a bias of 1 mpg, or 2 %.

Or, you coudl just say that in the long run, this should even out.

 

I have to say that the calculated LMPG calculated from total miles and total gallons used is approaching the average tank miles by the MFD. I would guess that these two will converge. I have a year and a half on the car and they are approaching each other as a limit.

 

if using just the MFD when taking tank readings, you maybe short. better to reset one of the trip meters when you fill your tank. that way you reduce your calculated error

 

Please see my post at other current MPG topic.

 

I've been wondering....

The energy monitor only shows a max 99.9 MPG. Does the Prius computer use this figure or the actual current MPG (if over 99.9) in calculating the overall average?

 

Actual. The scantool apparantly can show numbers up to 9999.99 (search 9999.99 to find the post i read this on)