Yeah... did you read the same article as I did by "click & clack"?

Canadian gas stations were the first ( and only? ) ones to put in pumps that based the volume on temperature. The pumps cost $3000 more, but Canadian stations were worried they would lose money selling gas that was colder than 60 degrees F. Most U.S. locations don't worry about that. Minnesota may be an exception! In December frost was 24" down... but I'm sure the tanks are very deep and never freeze, but may dip below 60 for a few weeks.

I reported here, a few weeks ago, that the local newspaper had an article saying that many of the gas stations in this area have failed weight-and-measure tests. I was getting a huge difference between the MPG numbers displayed by the car and the MPG numbers that I computed from the number-of-gallons displayed by the gas pump.

I have an update on that...

The gas station that I had been using was visited by the state weight-and-measure enforcement guy. I see the new stickers on the pumps. I also see more than half the pumps shut down with state seals on them, and yellow tape around them, and orange cones in front of them. Seems the pumps really were dispensing less fuel than they were charging for.

When I filled up at one of the other pumps, the calculated MPGs (from the pump gallons) agreed with the car displayed MPGs, within one percent.

My guess is that the state people took exception to the newspaper report and actually started enforcing the law. I don't expect them to keep it up, but it is welcome while it lasts.

I suspect the MPGs displayed by the cars may actually be accurate. At least, I think mine is. In the future, when a pump tells me a number-of-gallons that calculates a very low MPG figure, I plan on complaining to the state regulator. In Florida, the contact to file a complaint is at http://www.doacs.state.fl.us/standar...o/contact.html
I expect most states have contact information for their weight-and-measure regulator on the web, if you go looking for it.

After about 150 tanks in my HCHI, I find that the dash gauge is about 7% high: when it says 53.7 it calculates out (miles driven on tank divided by gallons pumped) to 50.0 mpg, on average. I say "on average" because the variation has been everywhere from 0% to 15% depending on whether I have underfilled or overfilled the tank. And that depends on whether the car is sitting level at the pump, or whether the pump clicks off early or late. I get a pretty close estimate of underfill by watching for the fuel gauge to drop from 20 bars (full) to 19. In my car, that point is reached when I have used 2 gal. from the tank: when I have driven, say, 100 mi with an indicated mpg of 53 or so. If I have gone only 25 mi when the gauge drops to 19 bars I know I must have underfilled by about 1.5 gal.

Taking account of underfill and overfill, my experience is that nearly all pumps I've used in the US are accurate to within 1% or better. Not the case in Mexico, where I have driven extensively on a number of occasions. In Mexico the pumps are as much as 10% off (guess which way!).

Overfill can be as high as 2 gal or more - I have sometimes driven 200 mi before the gauge drops to 19 bars. My experience is only with the HCH, but the underfill/overfill issue probably comes up in all cars, so if you get an indicated mileage that is way off from what you calculate, don't immediately blame the gas station.

I recently ran a complete TPC on my Ford's fuel system and it gave a 41HPE before fueling but after refueling and squeezing extra gas in the tank it came up as H24TL. I don't know if this is simply a calibration error or if Ford modified the mainline fuel return operator. The only thing I know for sure is that if one were to consistently fuel in such a fashion, the injection sequence would go disfarp and the repair would cost thousands.