About gas mileage.

Their reply seems a little simplistic. I can certainly see why the error is high for a single 10-gal tank of gas, but over 100 or 500 gallons, the error would go away. Some fuel stops would get a little more gas, some a little less gas. I think the Prius computer just overestimates gas mileage. Over a lot of miles, you can't get more accurate mileage than miles divided by gallons. Fuel pumps are accurate, I think. Lots of incentive to not give you more than you pay for, and regulations and inspections to ensure they don't give you less.

 

I agree. Their analysis is not very good here. You can't take one tank and use that to assess the trip computer, but over thousands of miles you can get a good read from it. Take a look at my Fuelly. For each Prius tank I record the MPG according to the car. It's consistently 6-8% off. For the Fusion Hybrid I record the gallons according to the car. It's also consistently a little optimisitic.

 

Why do feel that the error will go away after a lot of fill-ups? Won't the same apply to the Prius gauge? If you cannot refill to within 96-97%, that is the error rate most quoted for the gauge reading, when, in fact, it is the refilling inaccuracies that are causing it. My ScanGauge says that the car's reading is pretty good.

 

I believe that the error goes away because I think that the odometer is accurate (aren't there regulations about that?) and the fuel pump is accurate (I know there are regulations about that). The larger your numbers, the more the small errors at the pump and with you odometer will become trivial. Measuring flow of gas at any given moment and extrapolating miles/gal is going to be inaccurate because you're talking about small numbers with inherent, though small, errors.

If I drive 5000 miles and put 100 gallons of gas in my tank, I am getting 50 mpg because that's how the unit is defined. If I drive 500 miles and put 10.1 gallons in the tank one day, I can calculate my mileage at 49.5 mpg. If the next time, I drive 500 miles and fill the tank using 9.9 gal, my calculated mileage is 50.5 mpg.

Assume my mileage is exactly 50 mpg all the time. Also assume my "true" tank capacity is exactly 11 gal, but I fill up the tank when it can hold exactly 10 gal more (so when I've driven exactly 500 miles), with some amount of empty space in the tank, like 0.5 gal, and some reserve gas, like 0.5 gal. That means I can overfill a little or underfill a little every time I get gas. I can't fill the tank with 10.1 gallons every time because it will use up that extra 0.5 gal of empty space and won't fit. Nor can I fill it with 9.9 gallons every time because I'd eventually use up the extra 0.5 gallon and run out of gas.

I'll bet that was more explanation than you wanted.

 

After a few tanks (say 3000 miles) refilling inaccuracies have no influence when the cumulative amounts are used. My long term results say the HSI was about 3 -4 mpg high my previous Prius 3 driven 5600 miles.

 

It's SIMPLE math: The BIGGER the numerator and denominator become, the LESS variation there is in the final QUOTIENT answer!

 

As Barbie said, "Math is hard!"

Anyway, here is a snapshot of my Prius data for 2014. The displayed MPG is always higher than calculated, and the variance between the two? Well, it varies.

But it's the last column that tells the tale - It's the overall MPG average, starting from July 2009. While it appears to be dropping, it is incorporating seasonal fluctuations. It has hovered between 47-49 MPG over the past four years.

 

I've tracked every tank for over 3 years, virtually since the car was new. The dash display of mpg, well liters per 100 kilometer in my case, is very consistently positive. It's never been off the other way. My current average error, taking into consideration all the calculations, is 7.67%.

(This looks to be worse than RRXing's 2010 IV, above. The build date of mine was one of the first I think, summer of 2009. It's a Canadian Touring model, with the 17" wheels.)

 

My build date was May 2009. Purchased 07/08/09. Tracking all data since then.

 

Checked: build date says 08/09. August of 2009 I think. We purchased in November 2010.

 

Nice spreadsheet layout. Reminds me of what I *HAD* for our defunct 2011 LTZ Cruze, and am now re-instating for our new 2014 Prius (but only two fillups so far).

 

Thanks. Just as a visual reminder, the blue-shaded columns have formulas in them.

 

Do you do any "insight" calculations, like "change-in-MPG" vs. "average ambient air temperature"? I did and it was interesting to be able to anticipate the Fall-season drop in MPG ahead of it actually occuring after the first year of ownership.

 

I thought about using average air temp, but it's easier to just look at this chart (the anomaly on 11/11/12 was due to a partial fill-up following Hurricane Sandy.) Obvious mpg increases/decreases due to the season, with this past winter being exceptionally cold.

 

Looks like you have more seasonal changes (plural) then we do here in AZ.

My graph was lowest in Winter then rose to a 'peak' at the end of Spring, then plummeted through the Summer months (lots of A/C use), rose again through the Fall, then sank again to Winter's low. I was usually able to correlate the change from Winter-RVP gasoline in Spring, but not during Fall...weather changes swamped anything that I could detect.

 

I've tracked my mileage for over 90,000 miles (and set up the spreadsheet in such a way that I could also calculate the lifetime "computer displayed" MPG as well as the hand calculated MPG). My car has shown that the displayed MPG is 5.7% higher than the hand calculated mileage (63.7 mpg vs 60.2 mpg lifetime....soon to be 60.3 mpg). After 90,000+ miles, any variations due to tank fill ups, etc, have been eliminated or are so small as to not affect the overall numbers.

The displayed value appears to have a determinate error in it.

 

Yep, sounds like just about every manufacturer's ECU-calculated MPG values are between 5% and 7% optimistic.