Ethanol & capacitance fuel-level sensors

So, Jim Weir has posted some time ago in one of his columns, a design
and instructions for home-building a reliable fuel-level sensor which
uses capacitance and no moving parts to measure fuel quantity.

And many hombuilt aircraft use auto engines that don't mind (some) ethanol.

And auto fuel containing an awful lot of ethanol is likely to become
commonplace in the future, with some ethanol being all but unavoidable.

BUT both the dielectric constant and energy content of ethanol are
substantially different from that of gasoline. So if we are using
ethanol fuel, we want very much to know not only how much is in there,
by volume, but also how potent is the brew?

Assuming one had an engine and fuel system that didn't mind large
percentages of ethanol (presumably adapted from some future auto engine)
would there be any substantial difficulty involved in setting up an
in-tank capacitance system that measures ethanol percentage, and fuel
level, at the same time?

The system illustrated by Mr. Weir uses a capacitor formed between a
large plate and the (presumably aluminum) tank wall, with total
capactiance which varies depending on how much of the volume between the
plates is filled with fuel, and how much with air. What I have in mind
is to add to that a small plate at the bottom of the tank, presumably
always fully wet, to measure the dielectric constant of the fuel, and
automatic recalibration of the main fuel-level sensor depending on the
capacitance of the ethanol-percentage sensor.

I know this can work in principle, but what I don't know is what kind of
electronics are required to make it happen, and whether it can be done
on the same dead-simple scale as Mr. Weirs fuel sender, or would require
calibration to some nonlinear curve. (is the change in dielectric
contant as the fuel goes from all gas to all ethanol linear?)

Does anybody have any thoughts on this?
Does the mixture stay mixed, or does the ethanol try to float to the top?
Can we measure water in the ethanol, or does that just foul up the
sensing completely?

~Adam

"Adam Aulick" wrote in message
t...

Does anybody have any thoughts on this?
Does the mixture stay mixed, or does the ethanol try to float to the top?
Can we measure water in the ethanol, or does that just foul up the sensing
completely?

~Adam

Percent Ethanol sensors were once common in the E85 capable automobiles
(early '80's?). I don't remember how they worked though. They have since
been cost reduced out of the system in favor of estmating the ethanol
content by the magnitude of the closed loop fuel correction made via the
exhaust O2 sensor.

The stuff usually stays mixed.

--
Geoff
The Sea Hawk at Wow Way d0t Com
remove spaces and make the obvious substitutions to reply by mail
When immigration is outlawed, only outlaws will immigrate.

When marriage is outlawed, only outlaws will have inlaws.

Jim

When immigration is outlawed, only outlaws will immigrate.

RST Engineering wrote:
When marriage is outlawed, only outlaws will have inlaws.

It's very satisfying when your thread attracts responses from a
knowledgable source... ;-

~Adam

The problem with ethanol in aviation fuel is separation.

Ethanol in gasoline works fine, so long as the materials in your fuel
system and engine are made to tollerate it. But we treat avgas and
automobile gas very differently. Auto gas rarely sits for weeks
undisturbed, and fuel filler caps don't point up.

If even a small amount of water gets in to a blended gasoline, it mixes
with the ethanol and causes it fall out of suspension in the gas. It
then form a water/ethanol layer on the bottom of the tank, that may or
may not be burnable. It certainly won't burn the same way the gas
would.

In that condition, with a separate water/alcohol layer on the bottom, I
can't imaging the fuel sensors would read normally.

Once the alcohol falls out, the rest of the gas may be burnable, but
it's octane is significantly lowered. The ethanol is used as an
anti-knock agent. So you may get detonation problems. The fuel
chemists I've talked with say the raw feed stock gas (before the
Ethonol is added) is in the range of 84-86 octane.