# How to fill gas tank for MPG calculations

I understand that the most accurate computation of miles per gallon is over a long period of time, multiple fill-ups. But sometimes I want to determine if a particular driving technique would have a significant impact on my mileage. Like driving from San Antonio to Dallas, on the freeway at 55 miles per hour all the way (on cruise control). To do this, I would fill the tank at the beginning and then fill the tank at the end and use the mileage and amount of gas purchased.

But this depends upon being able to “fill the gas tank” to the same level each time. If I fill it more to start than when I finish, I overestimate my mileage and if I fill it more at the end, I underestimate.

I can’t “see” the actual level in the tank, so I need some other method.

Is the automatic cut-off by the pump accurate enough, on different pumps, that it will always stop at the same level in the tank? Or could it vary by enough to make this an inaccurate method of making sure that the tank at the beginning and the end is filled to the same point?

jim

If you can fill it at the same gas station and pump until it clicks off…that’s about as accurate as you’re going to get.

I use the same station and the same pump. Jot down the milage the first time you fill up. The second time you jot down the milage and gallons. Subtract the milage and divide by the gallons, which will give you the MPG.

It’s not always going to be exact, but it won’t always be far off either, and there is no other method. Some pumps will be different by half a gallon. This isn’t that big of a deal; you have to keep it in perspective.

Potential variables are not only the pump shutoff sensitivity, there is also the parking pad levelness, road surface condition, wind direction, curves in the road, levelness of the terrain, air temperature, gasoline consistency among brands, and traffic conditions are a few that come to mind. What you want is not possible; too many other variables.

As a practical matter, do what you can but remain skeptical about your results until you can see a pattern.