We know in general gasoline floats on water.
However, does ALL of it float?
Or are there tiny, perhaps microscopic droplets of water suspended in the mix?
I guess you would have to consider how recently your gas tank was agitated.
( Driven, filled up, etc. )

I suspect you are trying to resolve ( measure ) very small differences.
Say a density of .6997 vs .6999 ( just made that up for example ).

You may think you had better fuel with the .6999, but not of that .0002 was from a tiny amount of contamination. ( does not have to be water )

 

bwilson4web — Those are indeed puzzling results! Is it possible that some of the gasoline you tested earlier was "winter" gasoline, whereas you're now testing "summer" gasoline? By the way, were both Shell gasolines 87 octane?

I also find it strange that in your second graph (density vs temperature) there are lines of two distinct slopes, and the two Shells have different slopes. I can believe that different octanes might have different slopes, but the same octane? I have a possible explanation, and if it's true it might be confounding your attempts to run down this gasoline performance thing. Shell informed me that, here in Canada, in order to meet the government's current requirement of 5% ethanol overall, they are putting 10% ethanol into their Bronze (87 octane) and Silver (89 octane) grades in the major centers only, and not putting any ethanol into these grades elsewhere in the country. Their V-Power (91 octane) contains no ethanol anywhere in the country. [Here in Waterloo, Ontario, all the Shell gasolines are free of ethanol.] If some of your gasoline samples contain more ethanol than others, or if the ethanol content varies from month to month (or station to station), this might explain some of the results.

Stan

 

I work in the ethanol industry, and I will tell you the blending percent is only approximate, and not very scientific.

An example:
We will run the pump from the 100% ethanol tank at 10% speed, when running the pump from the 100% gas tank at 90% speed, into a "finished product tank" ( with additives also ) that is supposed to be 10% ethanol.

The end result is, the ethanol % will vary as much as +/- 2%, even though it is always marketed as 10%.
Or the reverse, E85 varies an equal amount, 2 or 3 % either way.

I was shocked how "low tech" this approach was when I started working there. But when you gotta move billions of gallons a year, I guess you can't be too picky.
-John

P.S. I think the different slopes could come from different additives, ethanol being just one of them. And in the U.S. I think it is legal to have up to 1.3 percent water.

 

The winter vs summer blend is a viable hypothesis but I won't be able to test that for another six months. One possibility is my earlier Shell 87 samples were their early adoption of 'summer' blend while the other bands were still using the less energetic 'winter' blend.

In engineering there is a certain amount of variance in our measurements. I used three temperature points to draw each line. To my eyes, the differences between the two Shell samples is small enough to be measurement error. The Cheveron is too close to call but the Exxon is distinct.

I'm only testing 87 octanes in this series.


Interesting. I haven't found any pumps in my neck of the woods with an ethanol stamp.

Bob Wilson