My sons truck did this and a guy changed the alcahol content. What usually goes wrong that makes them change?
As mentioned above, short fills and short drives are #1. The virtual alcohol content algorithm needs a minimum of 5 gallons during a fueling event, then it needs at least a 7 mile drive after. Inaccurate fuel gauges come into play.
After the above issues are addressed, then it comes down to intake air leaks or mis-read intake air. If the fueling conditions are met, on the 7 mile drive, the computer watches what the fuel trims do after the fuel fill. If the trims start to trend positive (adding fuel), the algorithm switches to the E85 table and sees if a point on that table starts stabilizing the trend. The point where the trims normalize (+/- 5% ish) the computer converts the map used into an alcohol percentage, stores that figure, and uses it to base fuel mapping until the next refuel event. If no map will correct the trims, then the ECM aborts the test and sets fuel trim codes.
So, it takes an understanding of what the system is doing, and why it is doing it. Any air leaks that introduce unmetered air, but not bad enough to set full blown fuel trim codes, will skew the alcohol learn. Intake gaskets leaks, stuck or dirty PCV, cracked air intake boots, etc. Exhaust leaks before O2 sensors can also induce this because leaks also let in outside air that can skew O2 readings. Other things that can cause this are dirty or inaccurate MAF sensors, missing or chewed up air filters, and aftermarket air intakes.
The diagnosis is straight forward. Test the actual alcohol content in the tank, reset the alcohol percentage, and test drive. Immediately after reset, monitor real time fuel trim. If the trims are higher than +5%, but less than +25%, you likely have an air leak or MAF misreading incoming air. For trucks, I come to a dead stop, WOT, at the 1-2 shift, the MAF should peak above 200g/sec. (5.3 V8s). If it is under 200, ypu have a dirty or bad MAF, or an air leak.