loose gas cap v mpg display
 

I forgot to put the gas cap on after a fillup once and drove about 50mi. I also noticed that my mpg went from 32[tight cas cap] to 50[loose]! Whaaaat? is that real? I put it back on tight, and drove 50m more. MPG went back down to 32. Then went another distance with it slightly loose, and got 48mpg on the display. So it is repeatable, but what is going on?
I think that the display is being fooled to give a high reading. The emissions system keeps a slight vacuum on the gas tank. When it does this with the cap off/loose some fuel in the form of vapor or liquid enters the engine and bypasses the fuel injectors. Since the fuel injectors are used to calculate the amount of fuel used, the denominator of miles/gallon is reduced and the displayed MPG is high.
When mpg is checked by [miles driven] / [gallons of gas pumped] it stays close to 32 previously observed.

Has anyone else observed this affect?

24 hyundai tucson hybrid

Is this calculated mileage by hand, or just what the display is reading? I'm thinking the system for the display is just thrown off by a loose cap. Unmetered air is entering the system, and is picking up vapors on the way, so the O2 sensor tells the computer to back off the fuel trim. You're still burning the gas, just some of it is bypassing the injectors so it isn't measured by the display.

the big variation is on the display. [I thought that would be clear from the shorter distances noted compared to 300+mi of a full tank]

looks like your in agreement with my hypothesis that fuel is bypassing the metered injectors. [but i don't think air is bypassing the metering device = mass air flow sensor]

still curious if others observe the same affect.

A loose cap is probably all that's causing the display system to malfunction. Because the air entering the system is not metered and is absorbing vapours, the oxygen sensor instructs the computer to reduce the fuel trim.

Last time I had a loose fuel cap was on a car without a mpg display. I knew it was loose cause it lead to the CEL coming on. When that happens, many cars go into default running modes. So the display here could just be making the calculations off some assumed numbers instead of using actual measurements from the engine.

Looks like most replies agree that the display is either going into a default mode or getting spoofed by fuel bypassing the measurement point.
Has anyone tried to duplicate the result? It would be good to know if its observed on other cars or brands.