MPG guage accuracy

Have you tried a search? This questions gets asked at least 3 times a week. I bet you find a ton of info.

 

There have been some 2010 posts as well. The 2010s lack the fuel bladder, so much of the uncertainty has been removed from manual calculations. Two or three mpg is in the normal range of error for most Prius. Whether the error is in the car computer or the gas station pumps is a matter of debate.

Tom

 

5% (3mpg @ 60mpg) is close enough for government work. IMO this illustrates that the bladder is not the sole cause. I guess we're still looking for the 1st car with an accurate fuel gauge.

 

I question whether it is always accurate using gallons used. When is the tank considered full? Is it full when it clicks off or when the gas is almost leaking out the top? It will take more gas if you fill it slowly vs fast. I think because of these factors, it is easy to be off 5% which is equivalent to about 1/2 gallon per fill up

 

Anyone know of a model/car the computes MPG that underestimates? I haven't, interesting they all overestimate...

 

Whether the fuel gauge is accurate or not won't have any impact on MPG actual vs. MFD. The gauge will show short term fill variability, but not an overall accounting deficit.

Some sort of negative departure should be expected in the MFD vs. real MPG even if the MFD calculation was theoretically perfect. There are several small sources that will result in some fuel being lost vs. what the MFD indicates has passed through the injectors.

 

:rofl: I doubt we will ever find a car that underestimates mpg's. I think that as long as mpg computation method is stated, we can have accurate comparison's between driver's.

 

It's very tough to know which is more accurate...but I'd say a computerized read out beats a fill up calculation any day.

 

Gas station pumps get certified regularly and to a very tight tolerance. So, unless station fraud is happening, the pump should be very accurate.

 

I think this may be it as on my last fill, slow, I still didn't see the gas in the neck after 3 clicks when I stopped. I also use a long leather string to see where the fuel is on the first click.

 

Is that your MPG for the 4300 miles? I gather that your car computer is reset at each fill up? It would be interesting how is would compare if not reset.

 

I agree. And if it was particular station fraud it would show up noticeably in fills at those locations.

Folks have spent too much time looking for conspiracies and invested too little time in examining real sources of error.

I expect there to be a negative departure from the MFD calculation. Why? Because the MFD only measures fuel entering via the injectors. It does not measure any fuel entering the combustion chambers via the vapor recovery system. It does not measure any fuel that is spilled or otherwise lost during the fill. So if the MFD was perfect at measuring the quantity injected, and the station pump was perfect at measuring gasoline dispensed, one would still anticipate that the MFD will indicate less fuel consumed than was dispensed.

And there are other sources of error such as differing temperatures. The density of gasoline does change with temperature (a cubic effect) while the critical injector dimensions would most likely change less and would result in an area scale difference (squared.) Plus the volume change of the liquid is likely to exceed that of the solid injector parts. So the reference temperature for the injectors could result in MFD calculations of excess total fuel consumption in summer and an undercounting actual fuel consumption in winter. I had some hints of this sort of impact this winter as my cumulative error jumped, but did not recover when warm weather resumed. I'm waiting to see if it happens again this coming winter.

And I'm sure that I'm just scratching the surface, and others probably could come up with better ideas. I expect that there are larger sources out there that I've not contemplated.

 

The saying is "garbage in, garbage out." The computerized read out calculates fuel consumption off of the injector pulses. It doesn't meter the fuel coming through the nozzle. Both could be perfect and still differ. On average I would put my money on the pump since it is upstream with more constant temperatures (underground tank) and has periodic calibration checks.

Now fill variability is substantial in any car (and more so in the GenII) so the MFD should be closer to the mark on each fill (precision) but the hand calc will be more accurate cumulatively.

 

I'm not sure I truly trust either the MFD or the pumps.

Pumps only get recertified every few years. In that time, they can develop discrepancies - deposits, wear and tear on moving parts, changes in weather (summer vs winter), changes in viscosity (summer blend vs winter blend), etc. The news often runs stories about pump accuracy and according to the shows I've watched, the pump will more often error in the favor of the customer (dispensing more fuel than the readout indicates).

The most of the same discrepancies are also found in the MFD. However, most cars never get certified, let alone recertified. I don't count the testing of 1 car for every 10,000 produced as a certification since that only certifies the one car tested, not the whole lot.

Lacking a need for a scientific result, I use a network time approach to gas mileage - I average all of the available sources together and claim that as the "right" value (pump and dash). I'm sure the hypermilers out there do things differently.

 

One idiosyncracy with the MPG number on the G2 Prius is that the number displayed is a continuous average. Right after one hits the RESET button, the MPG number changes avery 6 seconds. But as you accumulate miles, it seems to update slower and slower to the point where at the end of a full tank it seems to not be changing at all. This is due to its continuous average technique. Right after you hit the RESET button, the computer says, 'okay, you've gone a mile and burned nearly no gas so your MPG is 80.' Toward the end of the tank it says, 'okay, you've gone 300 miles and burned 7 gallons of gas so your MPG is 42.8.' See what I'm saying?

I suspect the G3 Prius works exactly the same. Right after you reset it it will show very high mileage and seem to update every 6 seconds, but toward the end of the tank it will show a 'true' MPG number lower than what you saw in the beginning and seem to update much slower.

That said, I would trust the computer MPG number more than a calculated number because the computer knows exactly how much fuel you've burned at any one moment and continuously keeps track of that number.

 

You are in California, I believe. This is from a previous thread on this topic:

http://priuschat.com/forums/gen-ii-prius-fuel-economy/21569-lousy-computer-readings-bogus-2.html

From yet another thread:

http://priuschat.com/forums/gen-ii-prius-fuel-economy/2005-inaccuracies-pump.html

As pointed out in the previous posts, this doesn't mean the MFD figures are accurate, but it does cast suspicion on gas station pumps.

Tom

 

I'm not so sure about that witht he GII. How are you determining the update frequency for the cumulative mpg later in the tank? Because you can't do it by waiting for the indicated mpg to change. The sensitivity of the decimal change in will be a hundred fold less even if the updating is unchanged.

I'm not saying that there is no change in the frequency of the cumulative MPG updates later in the tank, but I'm trying to understand how you are determining it.

This is not a continuous average as in moving average. It only requires measuring the ratio of two discrete sums. All it should have to do to update is take the difference between present mileage, and the mileage at the point of reset, and divide that by the sum of fuel injected since the reset. The gasoline sum would work as a digital totalizer since the last reset. It's not like it would make sense to do an integration of the area under the curve for the whole period, that would be wasted/unnecessary calculation.

The instantaneous display bars are different of course. They are time based and not accurate for determining actual MPG on a tank. (MPG doesn't need time as a factor at all.) They would heavily bias the MPG toward that acheived when idle or at low speed...as evidenced when one drives a few feet at 99.9 mpg, then sits a few minutes without the ICE cycling.

 

One way to know for sure is to call the county weights and measures department, or perhaps the county commerce if that is the one that certifies scales pumps, etc. Or, even the station should tell you how often it is certified.