The 4% is an overall average, some tanks are closer than others. One time the display showed 65.8 but at the pump the calculation was only 58. My HCH seems to display closest to reality when the mileage is low but those record breaking tanks seem to throw it way off.

We can test the idea next tank because the car is displaying a record 72.6 so far.

 

On my 2004 HCH, I keep Trip B as lifetime mpg. It currently reads 51.8 when my calculated mpg is 50.3. Again, that 51.8 could be anywhere from 51.5 to 52.0 (it would have to be closer to 51.5 because it just clicked from 51.4 during the last tank).
Therefore, the difference is about 1.5 mpg or just about 3 perenct optomistic.

 

update - my latest tank showed 51.8 mpg, but actual mileage was 50.9 mpg. only 0.9 mpg off. usually it is around 1 mpg optimistic for me, which is really not that bad.

 

My '03 HCH was off 9.3% over the 32705 miles driven. 'At the pump' mileage was lower than the trip computer. I kept track of all of my tanks. I just got used to it, and knew how much it was off. It hurt to see that '03 HCH go, but I had to trade it... just can't fit 3 carseats in the back of it.

On a positive note... my '05 Accord Hybrid DOES allow three carseats PLUS it looks like the trip computer (and NAVI) are near what I figure 'at the pump'. I'll see if that continues as I put a few tanks on...

 

I have a 2004 HCH CVT AT-PZEV, bought and used in California.
I fill up within 20 miles of when the refill light comes on.
I do not top off; when the pump stops itself that’s it.
I turn the cap at lease 3 clicks so it seals well so there is no evaporation.
I keep my tires at 40 psi.

Here is the % of error for my 25 tanks:
2.25
2.25
3.27
5.7
1.9
4.26
2.6
4.8
3.81
5.26
1.49
4.63
2.2
5.44
4.22
2.38
5.4
3.47
4.98

The average is 3.7 % (2.2 MPG) higher than what I calculate at the pump.


I suspect Honda is accurate and all the gas pumps are calibrated on the stingy side of the spec.

Do the Prius owners have the same complaint?



Hey, could higher tire pressure be the source of the discrepency?
Higher pressure means a slightly higher tire diameter.
Right?
So one rotation of a larger tire would take you a longer distance.
Right?
This would translate into higher milage calculated.
Right?
Perhaps the HCH is calibrated for people who keep their tires at 32 PSI.

Perhaps a test is in order.
I could lower my PSI to 32 for one tank and see if the error goes away.
But I really hate to lower my MPG.
Anyone else want to volunteer for the test?

 

I have a 2003 Civic, and it consistently reads 3 mpg higher than the hand calculation. I am sure the discrepancy is in the Honda rather than a nation-wide error in pump accuracy. The only place I have seen pump errors more than 0.1% is in Mexico, where they are nearly always about 10% off in favor of the station. It has nothing to do with topping off - my numbers are based on a lifetime A reading of 50.4 and a lifetime (over 32000 mi) hand calculated average of 47.2. By the way, the odd numbers displayed on the dash (what is a FPR?) come about because Honda calculates the mileage in miles per liter, rounded to the nearest 0.1 mpl, and then converts to mpg.

 

These pump discrempancies are occasionally published and you would be suprised how often they're off. And it is almost always in their favor....

 

"... could higher tire pressure be the source of the discrepency?"

That's an interesting theory, and one that I had not heard proposed before. As a data point, I keep my tires inflated at about 36-38psi, and my MPG is usually off by about 1-2% (0.5-1mpg).

 

but the car doesn't know your air pressure and uses the same calculation you would to get its numbers. Unless you are using some external way of measuring miles then what you measure by looking at the odometer and what the car measures is going to be the same no mater what tire size.

Sure, tire size makes a difference in what the odometer reads but its irrelivent because it is off for both you and the car for the calculations.

 

NOT if you have NAVI. The TC in NAVI is figured using GPS and has proven, by my calculations, EXTREMELY accurant.