Scanguage II fuel consumption calibration for E0 fuel

My Scangauge shows 0.02 GPH consumption rate when it is actually zero, with engine stopped, with car stopped or coasting at less than about 45mph. It sometimes incorrectly shows around 0.33 GPH when the engine is being turned by the wheels while coasting at more than about 45 mph. In the latter circumstance, at other seemingly random times it (correctly) shows 0.00 GPH. Why? The MPG numbers it shows are affected correspondingly, MPG being equal to MPH/GPH.

IF the ScanGauge uses the false ~0.33 GPH consumption rate integrated over time to compute fuel consumed for calibration purposes, the calibration will be way out of whack. That said, my apparent correction factor is around 12% on E10. It seems inconsistent, possibly because of the 0.33 GPH problem. (The 0.02 GPH issue would throw off the calibration too, to a smaller extent.)

The only way I can confirm the engine is not actually consuming fuel during higher speed coasting is to note that the coolant temperature drops at about the same rate whether ~0.33 GPH or 0.00 GPH is displayed.

 

Actually the 2.1 gallons after flash works for the car sitting in my driveway. I would expect that other Gen-3 Prius tested the same way would have a different results:

1. When flash begins:
   1. reset a trip meter
   2. put exactly 1 gallon measured, spare gas can in the car
   3. drive routes where you can coast to a safe shoulder when the engine runs out
2. When car runs out of gas:
   1. add 1 gallon
   2. take photos of all trip meters
3. Drive directly to nearest gas station and fill up
4. Do the math and share with the community

Bob Wilson

 

This seems like a cumbersome way to determine correction factor. Why not hand calculate MPG and compare it to what the SG thinks it is, then input the correction factor?

 

Yes, and more cumbersome than the method ScanGauge's instructions recommend, by an even wider margin. That is to compare fuel burned between fill-ups according to the pumps to fuel burned over the same interval in the ScanGauge's estimation, and correct accordingly. The trick to getting a stable number is to do it over a long enough interval that fill inconsistencies are diluted, and make only small adjustments.