Computer estimate way off... 52.5 actual vs. 45.4 displayed

[Escalade_on_22s]

I just filled up for the first time this morning. The car took 9.7 gallons for 510 miles. That's about 52.5 actual mpg, but the display calculated a mere 45.4. I knew I would beat 45, but 7 pts seems substantial, especially since other owners seem to get a few pts less than what's displayed.

Driving conditions: 80% highway driving in mountainous terrain @ 65-70 mph, the remainder being short (<5 min) trips to work and back (which is where my averages start to fall off). I imagine the short commutes lead to inaccuracies in the system, as it appears to get about 20-30 mpg in the first 5 minutes regardless.

I understand the car isn't quite as efficient in the first few stages of operation after start-up, but I think 25 mpg for 5-10 minutes is way too conservative. It's probably closer to 40-50, even in the first few minutes of operation. Maybe a software upgrade is in the works. Or possibly the thing just didn't get calibrated until half way through tank 1, any ideas?

[olends]

with the flexible bladder in the tank. the amount that you filled, may not be the same as the amount that the car used. these kind of errors usual normalize over many tanks.

right around 25mpg at start seems correct. with the engine warming is loses its ability to stop to preserve fuel. if the car get 50mpg with the engine being off 1/2 the time it would make sense with the engine running 100% of the time the mpg would be half normal. 50/2= 25.

[efusco]

Just close your eyes and don't look at your mileage numbers for the first 4-5 tanks or so...then ignore any differences b/w the computer and fill mileage on any individual tank.

You have no idea how full the tank was when it was first filled or if the MFD had been reset. A different pump was uses that may have had an earlier or later shut-off. You may have underfilled the tank a little. The bladder size may have been smaller when you filled than when it was orignally filled.

Am I getting my point across? If I were a betting man...and I am...I would bet that your computer's mileage is closer than your calculated mileage. There's not really any positive way to know, of course, but time will be telling. Just keep tracking them for the first 10 or 20 fill-ups--until you're satisfied that the numbers the computer gets are accurate or until you can see how far off it actually is from your long term calculations...then you can adjust for it later on.

[mtc1234]

I just had my 2nd fill up, and had a similar discrepancy. 47.3 MFD mpg, 471 miles driven. Calculated mpg 54. In an earlier post, I suggested tracking/comparing the amount of fuel to fill up against the "calculated fuel burned" (miles driven / MFD mpg). For this go around it was 8.738 pumped versus 9.96 calculated. My first fillup was 9.414 pumped versus 8.58 calculated. Even taking into account the flexible bladder and vagaries of pumping gas, this seems odd to me. I fully expect the issues are how I pump.

But, what bothers me the most is that on the first tank, I filled up when I hit one bar. On this tank, I "pushed" it, having the blinking bar for 20 some miles. Yes, it filled up with a lot less gas! I have to figure out how to fill up in a way that I can trust I'm being consistent. In other vehicles, I've always been able to hear the gas going into the tank and could hear it getting close to the top. With the Prius, I hear nothing.

How *does* the MFD calculate mpg? Does anyone know? Should we expect it to be accurate and our calculations to merge to it? People talk about it averaging out in the long run - that implies that either the MFD calculations change (or some variable it uses averages out over time) or our own calculations change (or our habits of pumping).

All of these threads about mpg and differences b/w MFD and calculated seem like a lot of hand waving. Not that I care too much, I know I'm getting great gas mileage. It's just the engineer/scientist in me that is unsatisfied.

mtc (sorry about the long, dangling thread)

[Escalade_on_22s]

I agree, no one seems to know whether or not the MFD readout is based simply on computer modeling algorithms, or actually based on metering the fuel flowing through the injectors. I feel it probably uses a transfer function that takes into account mass air flow into the ICE and current speed (much like the initial mass air metering techniques used in early Ford EEC-IV of the mid 80's). I doubt you could meter fuel accurately enough to get MFD read-outs much more accurate than about +- 10-15%.

I plan on researching this and finding the answer. I don't think there's any possible way that fuel bladders can shrink over 1.5 gallons. For know I plan on figuring my MPG the old fashioned way.

[mtc1234]

I see in the other related thread that some good information is coming out. For my particular situation, I've discovered that I did indeed underfill the tank. My first two tanks didn't start dropping on the gauge until 150 miles or so. Yesterday, after 50 miles, it dropped a "box". I have to figure out how I can avoid underfilling.

[efusco]

MPG is calculated, by the computer, by counting the amount of fuel through the #1 injector... It tends to be very accurate.

[rflagg]

I think you'd be hard pressed to find any of us 20k plus drivers with a discrepancy between computed and calculated MPG with more than 2%. Per tank, ignore your own calculations. Trust the car.

-m.

[DaveinOlyWA]

i would definitely track both data methods for comparison purposes if any. you can see by my signature that they are slowly converging and are now nearly identical.

something i stopped doing almost immediately and might help you (although i didnt notice a difference) is that i dont top off my tank. i stop at first click, round off the dollar amount and call it good. there was a rumour (i call it that because no gas station owner will verify this that i have asked) that the overflow detector may drain gas back during auto shut off. granted the tube is very small but the more times you dribble gas in after shutoff the more potential you have to lose some of the gas you paid for.

[Prius Maximus]

Trust the car's displayed mpg. The bladder is too variable for guessing how much gas was actually pumped in or used.

I have over 36,000 miles. I have done 92 fillups. My calculated mpgs have ranged from 7.5 mpg higher than the car's display to 19.1 mpg below the car's display.

Overall, the difference between the calculated mpg and the car's display mpg is a whopping 0.1 mpg.

Go with the display value, but track you're calculated value for fun.