Fuel Tanks and ethanol....

[CDK]

Thanks for your effort CDN-CUCV.

The 1th mechanism cannot be present in an Al tank. As you stated in your opening line, the oxide layer (Al2O3) completely protects the metal so there is no further corrosion, even if water were present. The bond between metal and its oxide is very strong; in fact that is the reason why Al is a useful construction material.

To make Al2O3 in its pure form - a white powder - aluminum must be ground into very small particles so all metal is converted into its oxide and there is no bond left.

The presence of chlorine is another matter entirely. In water it ionizes and dissolves the Al2O3 and forms aluminum chloride which is water soluble. If enough water is present the whole tank will be eaten away.
But this takes place whenever seawater is allowed to enter the tank and is not related to the presence of ethanol.

The 2nd mechanism you mentioned is something I didn't think about: you may have a good point there. A requirement would be that the gasoline-ethanol mixture becomes conductive enough to act as an electrolyte.
I will do some quick tests with different mixtures and post the results here.

[gonzo]

I think the brass jets in the carb bowls and the steel shafts probably cause the electrolysis.

[CDK]

Originally Posted by gonzo

I think the brass jets in the carb bowls and the steel shafts probably cause the electrolysis.

No they don't.

I did the following tests:
A whiskey glass, two new tungsten welding electrodes, 1" apart and a digital multimeter with range up to 200 M-ohm.

-Pure gasoline-98: infinite resistance.
-A mix with 10, 25 and 50% ethanol (95% pure): infinite resistance.
-Only ethanol: 80-90 M-ohm.
-Ethanol with 60% water (also known as Wodka): 1.8 M-ohm.
-Same, but a knife point NaCl (salt) added: 260 Ohm.

Conclusion:

Adding ethanol to gasoline in any quantity cannot cause electrolytic corrosion, not even when pure ethanol with 60% water is used. Only salt can reduce the resistance enough to start electrolysis.

The amount of water in a gasoline-ethanol mixture must not be overstated. In a 50-50 mixture is was possible to add approx 3 vol.% of water, more was separated immediately after stirring. With 10-90 like available in the US, that would mean 0.6% if you try very hard.

Water in fuel can corrode the shafts in the throttles, especially if that is sea water.

[Bigfoot1]

I would contradict your comment that there is galvanic corrosion caused by either gasoline, or enthonal water in an aluminum tank.
Galvanic corrosion is the result of two dissimilar metals in contact with each other in an electrolyte, ie a free ion liquid, like salt water,
Your comment that this could happen the same as trim tabs as trim tabs are in sea water, an electrolyte, so you can get GALVANIC corrosion here, same with stern drives which will have dissimilar metals throughout the assembly is not correct. Corrosion yes, but not galvanic corrosion
If the boat tank is built and installed to ABYC standards, the aluminum tank will not be in contact with any other dissimilar metal that is far apart on the nobility scale to cause significant galvanic corrosion and certainly it would not be in contact with salt water to create this type of corrosion.

Your comment that you have had an explosion in an aluminum tank due to hydrogen build up. Where would the hydrogen come from if the tanks were flushed? Hydrogen cannot, nor other gases, permeate aluminum and secrete out with time, cannot happen,
We have repaired probably about 100 tanks in the last 30 years of building both steel and aluminum tanks, fuel tanks, gas and diesel, we steam them, then weld any areas that need it and have never had an issue.
So your statement that there was a hydrogen build up seem erroneous.
I would have to suspect that some hydrocarbon fuel was available as the source of an explosive vapor.

[CDN-CUCV]

I'm only basing my statement on the hydrogen gas from what the shop owner had told me. Apparently, pink bilge cleaner may have seeped into cracking along welds, cold-cracking , stress cracks I don't know. The tank was a buoyancy tank and not for fuel and it was possible that flushing didn't reach that space so hydrogen may have still been present. I'm not writing off hydrocarbons though, that to me makes more sense considering it was a gas-powered, open boat. I could have explained that better. apologies...

Regardless, boat exploded through the hull and started a fire which caused extensive damage to the shop.

you're are right about the galvanic corrosion. Freudian slip, my bad. I guess electrolytic corrosion would be the correct term perhaps?

[Frosty]

The quote earlier that brass or bronze plugs should not be used in Alluminium but instead stainless is surely the wrong way round?

I cant see anyone fitting ally to ally plugs with out teflon tape or some sealant which would stop galling.

[CDN-CUCV]

PVC caps plugs can work. I've seen brass as well with a little liquid or tape teflon, no big issues. Correct me if I'm wrong

[Bigfoot1]

the teflon tape will not stay in place on the thread faces, The idea around tape sealants is to seal the voids that are at the locatioin of the very tip of each thread. Pipe threads as well as parallel threads are machined to percentage parameters, ie not really tolerance but a designed clearance between the tips to the roots of the opposing thread. So when you tape a joint, it appears that the thread face is covered, and it is until you tighten the joint at which time the plastic deforms into the available space and leaves the thread face, try it sometime, tape a pipe joint, tighten it and untighten it and see where the teflon tape is. My statement was to use an aluminum compatible ( not copper based) thread sealant for lubrication if you were going aluminum to aluminum joint but better to use a stainless pipe fitting that when tightened into the tank is going to be attached to the tank pretty much forever. and then use a cap, or locking valve than can be removed. Of course, stainless on stainless can also gall if over tightened

Someone mentioned PVC caps, not legal by ABYC rules, cracking, embrittlement over time etc.

Frosty, 300 series stainless or aluminum fittings are the only approved fittings for marine fuel tanks in Canada and the US and one would assume the world as galvanic corrosion is not constrained by international boundaries
Just trying to be helpful,

[gonzo]

The brass jets in a carburator are in direct contact with each other. The test is valid for other purposes.

[Accurate twrs]

I like the idea of the sump with a tube at the bottom of it, fitting on top for vacume. Occasionaly i have customers who ask to put a drain valve on the bottom of the tank to which i say NO...!!! In the fitting department i use brass all the time on top of the tank not submersed with no problems at all.. On occasion i will make a brass and copper pick up tube thas's completely removable from the tank. Never had any problems. Anybody know what tank cleaner was used that took the life of the welder..?

[Bigfoot1]

while you may not have had problems, why risk the liability with using brass fittings in an aluminum tank. a small leak in a GAS tank and you can have problems. If ABYC say that brass fitting are not allowed, then why use them.
why???

[Accurate twrs]

Bigfoot... I've been useing brass and copper for nearly 25 yrs now and have never had cause to change. Never have i had a leak in my tanks either.. however I make sure my customers use the ground tab i weld on the tank...
I was surprised how many customers had no ground on their tanks.. Many a customer brought in tanks looking for just a repair, rather then replacement. I would refuse to repair them for the liability. As in no way Jose, not
when my tail is on the line. Currently im chargeing $9.50 a gallon. A certified tank is more. They buy the sending unit if they want one. If a tank has many facets there's a degree of diffaculty charge too.
If i'm missing something here let me know about the copper and brass...
thank you tom

[CDN-CUCV]

Originally Posted by Accurate twrs

Anybody know what tank cleaner was used that took the life of the welder..?

Tank cleaner didn't take the life of the welder. A pink bilge cleaner being was suspect in an explosion that started a shop fire, may have been hydrocarbons too, just going on what the shop owner told me at the time.

The fatality that occured, was another incident where there was welding being done on a boat and the fuel tank exploded. There was a plaque dedicated to that welder at my school.

[PDQAltair]

I have found this EPA memo enlightening. It explains the chemistry of phase separation.

http://www.epa.gov/otaq/regs/fuels/rfg/waterphs.pdf

It makes 2 things clear:
* That a drop in temperature is the main trigger for water drop out.
* The EPA was well aware of the problem.

I always had trouble in the fall, with the first cold snap. Without re-telling a long story...
* Full tanks prevent breathing. Fill after sail.
* Close the vents on small outboard tanks, if provided only.
* Small fill leaks cause BIG problems. ANY deck leak is fatal, since it takes a lot of ethanol to the bottom with it.
* Tanks that are protected from the sun or that are in a cool space have many fewer problems.

This also explains why some say "no problem" and other say "the sky is falling." They may both be correct. It depends on the system design.

[TollyWally]

"The EPA was well aware of the problem."

And the reason those bastards shouldn't beheld accountable?