how low can you go... fuel tank capacity

The condensation argument doesn't sound right to me.

If you always keep your tank full, lots of junk from the fuel can settle in the sump because its never getting close to empty, then the first time you have to drive until its near empty, that's when all of the garbage hits your fuel filter all at once.

I guess we don't hear about water in gas because almost everyone has an ethanol blend and for Californians, we don't have a lot of humidity.

With an Ethanol blend, if it can absorb water, going until near empty is going to remove most of the ~contaminated gas and then the fill up dilutes it to a great extent. If you never run close to empty, the dilution is not much.

For arguments sake, let's say 100% humidity is the same as straight water vapor which I think would weigh roughly 36 grams for 22.4 liters and our gas tank holds about 50 liters. Let's say we run it down to 5 liters gas and 45 liters of water vapor. That's about 72 grams of water vapor maximum or 72 cc's volume if it all condenses but you still have 5,000 cc's of gas. Worst case you're roughly 98.5% gasoline, 1.5% water. Real world its probably more like 0.15% because the tank would never be full of water vapor, then you fill it up with another ~45 liters of gas and the dilution takes the water concentration to 0.015%.

In regular conditions, I wouldn't worry about condensation unless regular conditions means driving 1 mile every day with big temperature swings to form condensation and fill ups so infrequent that water can accumulate.

In that case, I would want to use fuel stabilizers more than alcohol to absorb water or keeping the tank full.

That's funny, why do you think drag racers put extended sumps in their gas tanks?

Stock tanks pick up near the bottom, not the bottom. Manufacturers are funny about not wanting anything sticking out under the tank that could get torn off in a crash, especially Ford (ever heard of a Pinto?).

My condensation estimate is a worst possible condition just to find out what ballpark we are in.

So where are the conservative and scientific figures showing how keeping your tank ~full reduces water contamination? I didn't see anything accept a note that surface area is bad.

Give me a citation or show your work, they did neither.

Let's see your math.

For all I know, I could be wrong or you could just be calling me names.

I'm not afraid to be wrong, I just won't take your word for it.

I await a rebuttal and I may even look up a better estimate of how much water can condense out of 45 liters of air at 100% humidity instead of just guessing based upon my observation of fog and the density of O2, N2, CO2, etc.

To improve the calculation of the maximum amount of water available to condense over one tank of gas:

Assume a range of 330 miles on an 11 gallon fill up and driving 33 trips of 10 miles each, a temperature of 50 C (122 f) and 100% humidity.

After the first trip you have ~97% gas and 3% air. After the last trip let's call it 0% gas for simplicities sake. The average volume of air in the tank is 50% and you have 33 cycles so the maximum theoretical amount of humid air that can get inside the tank over that length of time is 181.5 gallons which is roughly 33 moles and multiplied by the molecular mass of O2, 1056 grams of humid aid can be in the tank. According to a relative humidity chart I saw, 100% humid air at 50 C can hold just under 100 grams of water per kilogram of air.

That means 100 grams is the max which won't happen in reality because gas tanks have whole systems designed to control evaporation so the air doesn't change after each trip.

My rough estimate is that 10 grams would be an incredible amount because the gas cap is designed to hold pressure and even a little vacuum I think. There might never be more than 11 gallons of air entering the system between fill ups (and that could contain roughly 3 grams of water).

If the gas is absorbing the water over that time and we are burning the gas, some of the water is going out with the gas so at the end of the test, there is probably less than 5 grams of condensed water contaminating the gallon or so left in the tank. My estimate of 0.15% water contamination due to condensation is looking right on.

What can I say, some times I am a good estimator.

I think that's a pretty normal use case, it matches my commute beside the humidity. An unusual use case with very short trips, 100% humidity and huge temperature swings could be worse if the gas tank is breathing a lot but your gas would be evaporating away too.

If anyone can improve on that, go for it. One place to improve could be EPA regulations for the gas cap system or the spec on the test smog stations do. Then you could do some PV=nrt to figure out how much air can get in over a big temperature swing.