Mileage discrepancy

I’ve started watching the fuel mileage in my car. Here’s what I’ve got:

Date	      Odo Miles	Gallons	Ind MPG	Calc MPG  Error
05/12/08	216.5	4.658	  n/a	  46.48	  n/a
05/15/08	207.6	4.691	  46.8	  44.25	  5.75%
05/31/08	300.3	6.864	  n/a	  43.75	  n/a
06/11/08	225.6	4.769	  50.7	  47.31	  7.18%
06/12/08	221.6	4.777	  50.2	  46.39	  8.22%
06/17/08	240.8	4.907	  49.5	  49.07	  0.87%
06/18/08	222.0	4.590	  50.1	  48.37	  3.59%

Odo miles is the number of miles indicated by the trip odometer.
Gallons is the number of gallons pumped, with no ‘extra clicks’ after the pump shuts off.
Indicated MPG is the number of miles per gallon indicated by the navigation system.
Calculated MPG is derived by dividing the ODO miles by Gallons.
Error is the percent difference between the indicated and calculated MPG [(Ind/Calc)-1].

The odometer on the navigation system generally displays a traveled distance of 0.3 to 1.0 mile less than the trip odometer. (It’s hard to tell, since the nav system odometer has no decimal places.)

Question: Why is the indicated MPG so much different from the calculated MPG? Should I trust the trip odometer, which I presume is electro-mechanically attached to the drive train? Or the navigation system MPG, which I presume is calculated using the GPS for miles driven? Why is there such a large variation in the %-error column?

Perhaps it’s a relic of a differing amount of gas entering a vehicle before the valve shuts off (which you did indeed try to control for?) Just .25 gallon difference would cause a 5% error in your first example (and that wouldn’t be too much out of the question IMO: if one station has a high cutoff and the other has a low cutoff they would each be only .125 gallons away from the “ideal” cutoff and still produce that variance.)

I nearly always use the same pump, at the same station, and usually around the same time. I insert the nozzle, pull it back until it stops, and then set the lock on the trigger while I go in to get my coffee.

But I suppose there are factors that would affect the shut-off.

Why wouldn’t the navigation be hooked up to your odo? After all, it’s hooked up to your fuel injectors/pump somehow (otherwise it wouldn’t know how much fuel you were using).

I wonder if the discrepancy can be explained by the reset. When I reset my average fuel economy, it always acts really funky for a minute or two until it settles in. I wonder if that has anything to do with it?

I don’t think the shutoff really comes into play, though, because the indicated mileage is always larger (from our admittedly small sample size).

As an aside, when my car determines distance to empty, the “estimated distance is determined by a weighted average of the instantaneous and average fuel economy, according to the current fuel tank level”.

Not to be a busybody, but around here there have been a rash of incidents where gas was redirected to a different vehicle while the person paying for the gas had wandered off into the store or was otherwise not paying attention. That would really throw your calculation off, not to mention your fuel budget.

Coming from L.A., I’ve thought of that. But where I am now it’s not an issue. And it’s pretty much just the few regulars who come in at 0530.

You seem to be getting impressive mileage. What are you driving?

2005 Prius I bought from my former roommate.

It looks like Toyota Prius mileage.

I just took my Prius on its first road-tip.

I got still 46-47 MPG even at highway speeds, but, unfortunately, the car was in a flash flood that occurred overnight while I was sleeping in the hotel and is now at the dealer getting fixed up. The floor boards were completely full of water, and after opening the doors to let the water out, we still had to bail the rest of the water out before taking it to a car wash to wet-vac the rest.

I was able to drive it home and it continued to get 46-47 MPG! Everything still worked, but it has some error lights lit and it thought it was out of gas every 2 to 5 gallons.

The dealer just called to say it would be $1800.00+ to pull the seats and carpet, replace the carpet pad, clean and sanitize the carpet, check/replace some sensors and put everything back in. He said the hybrid battery is sealed and did not get wet.

I don’t think you have a sufficiently large dataset here. In both senses. Your mileage is very low - only 1500 miles total - and low between refuelling. Try going from a full tank to an empty tank (or when the low fuel indicator shows) a dozen times. Have a full petrol can in the car, of course.


Please don’t.

AAA is your friend

I do agree that more data points are needed.

The number of data will increase daily, of course. (Well, three days a week.) But I don’t know if I can make it to work and back home twice on one tank, and I don’t really want to carry fuel inside of the passenger/trunk area.

If I get over this laryngitis in time, I might make a trip down to Coos Bay in a couple of weeks. That should be about 21 gallons. The car supposedly has an 11.9 gallon tank, but when I refuel the gauge is reading under half a tank. (The owner’s manual warns against topping off the tank, as it may damage it somehow.) So I’m guessing a ‘full tank’ is 8 gallons. I’ll find out when and if I make the trip.

The amount of gasoline pumped into your car is not corrected for thermal expansion. You buy gasoline by the gallon, and burn it by the mole. For every 10 degree C increase in the temperature of gasoline you lose 1% of the fuel volume. If your car’s calculation is based on standard conditions rather than actual volume used that would create a systematic error if the temperature of the fuel pumped was higher than normal.

So the more accurate measurement is my calculation based on the miles traveled divided by the gallons pumped?

The reason I’m asking this question in the first place is that I’m very suspicious of odometers, and I wonder if the miles-driven indicated on the trip odometer or the navigation system is more accurate.

Why not? It’s SOP over here. Cars don’t explode like they do in Hollywood films.

Since the Prius doesn’t have a proper trunk (or boot, if you prefer) things have to be carried in the general passenger compartment. (I don’t consider the under-floor comparment in the rear to be a trunk.) Not worried about explosions, and I’ve carried gasoline inside of a car before. But I hate the fumes.

You’re refilling only 4 gallons or so. U.S. gallons at that. This is a third of a tank - full being 11.9 U.K. gallons. Give it a serious workout. And put the petrol can in the boot / trunk.

Ahh, good call, Bill Door.

Technically, that measurement is more accurate for what you’re paying to operate per mile, while the car’s indication is a more accurate indication of what you’re actually using, assuming all measurements are accurate.

Do you have any evidence that the nav system doesn’t use the odo to determine miles driven?

I’m not all that familiar with the Prius. Does it have a GPS based navigation system or not?

Well, it’s a hatchback, not a sedan. What do you call the space behind the rear seats? We still call that the boot.

Besides, if you get yourself a proper can, you won’t be troubled by fumes. Rubber seal, friction compression, cotter pin to prevent accidental release, and you’re set.
Works for me.