Using the ScanGauge to assess pump calibration

[brick]

It looks like I have come across a new and interesting use for the ScanGauge: tracking measurement variability from station to station. For the first part of the summer I was filling at two stations in particular: A Shell here in town, and another Shell in RI. I had the ScanGauge calibrated to increase its FE numbers by roughly 5% to agree with pump calculations, and it was getting to within 1/10th of a gallon every single time (generally staying on the pessimistic side.) In August I started filling at a Sunoco on my work commute because the prices tended to be ~10 cents cheaper per gallon and it is convenient. My first fill at that station resulted in a fillup that was at least a half gallon more than the ScanGauge told me to expect, and I attributed that to a different level of fill. No big deal, that should iron itself out with the next fill at the same pump. But along came my next fillup and again it gave me more fuel than the ScanGauge thought it should. At that point I decided that I must be doing something different and changed my ScanGuage to a ~2.5% correction factor instead. After all, it's not a perfect device. This has been consistent with several subsequent fills at the Sunoco.

But now somthing strange is happening. I went back to that Shell as a matter of convenience and guess what happened? The fuel volume registered by the pump was lower than what the ScanGauge expected. So now I have re-set it to the original 5% correction with the intent to fill at the same pump next time. If the ScanGauge agrees again, I'll be able to make a safe bet that the two stations are a few percent out of calibration with one another, with the Shell being in my favor. (Over several tanks, they should be roughly the same.) I'll have to look at the inspection label to see if it identifies acceptable tolerances to assess whether or not that Sunoco is at the edge of the range or just plain out of it. (I have a hard time believing that the Shell would allow pumps to run 3% in the customer's favor.)

What it boils down to is that the ScanGauge has the potential to protect the consumer by comparing supposed fuel consumed with fuel pumped. You can't make an assessment based on one fillup, but it could pay to pay attention over the course of several consecutive fills at the same station.

[krousdb]

I researched pump error and found that most states set the tolerances at +/- 1/2%. If there is an inspection seal on the pump, l it should be within those tolerances.

One thing to consider, the Sunoco's around here use 10% ethanol. That is why I never get gas at them. I don't know why that would make a difference, maybe the lower energy content is screwing up the SG.

[brick]

I hadn't considered that, but it could be. 3% would be right about the mileage hit that one would expect. Thing is, every single station in this state has stickers announcing "up to" 10% ethanol. Last I checked we had it in all our gas, as it's about the only oxygenate allowed by Connecticut environmental law. I also would have expected the ScanGauge to pick up on the extra fuel burned, but who knows? You could well be right.