Gas fumes? Tank leaks? Read this.

[Monkey Wrench]

maybe an intentionally slightly-leaky filler cap can solve a lot of these issues. Yes in a rollover it may drip a bit causing a fire after collision, and it may also circumvent proper vapor recollection as intended to pass enviro standards. the vapor canister was never used or needed on older cars built before such standards were in place.

is this mitsibushi valve to replace any 944 valve or only certain models? it sounds like a reasonable solution.

I remember back when we had emissions testing they would check our caps to be sure they sealed, often cars would fail due to an old cap not sealing. I think there might be some differences between caps and their specs between models.

i remember there being a lot of sealed oil drums around where I grew up, we used them to float docks and they would typically be making oil canning noises as they expended and contracted in the sun, just the empty barrels.

i bet there are a lot of older cars out there with leaky caps and they are never a problem and thus never diagnosed. then any excess fuel vapors can leave through the cap rather than pressurizing the tank.

I assume that as fuel is used it draws back through the carbon canister rather than creating a vacuum, recollecting some of the fumes that the tank may have expelled during warming weather.. and that's better for the environment. I've seen cars with rusted out charcoal canisters that obviously weren't' sealed anymore but never noted any issues caused by it.

[TheSecondChris]

A slightly leaky filler cap does solve a lot of problems, but it also lets fuel escape if you do any hard cornering. That's what I'm hoping to solve. The mitsu valve is for the early cars.

[originalowner]

The purpose of the swap is to reduce the possibility of fatigue cracking of the metal fuel tank, while still trying to be conscientious about hydrocarbon emissions. A "defective" fuel cap, "defective" valve, or a cracked fuel tank will leak gas fumes, even while the car is parked. There is no reason to do this on a 924/944 with a plastic gas tank.

[Rasta Monsta]

Modern cars will throw CEL if the cap (or any part of the evap system) leaks, because fuel vapor is the highest pollutant in a modern car.

[TheSecondChris]

Originally Posted by originalowner

The purpose of the swap is to reduce the possibility of fatigue cracking of the metal fuel tank, while still trying to be conscientious about hydrocarbon emissions. A "defective" fuel cap, "defective" valve, or a cracked fuel tank will leak gas fumes, even while the car is parked. There is no reason to do this on a 924/944 with a plastic gas tank.

In addition to preventing a cracked metal tank, the high pressure allowed by the stock valve was creating a hot start issue for me. I had been having an issue with a very rough idle after a hot start, which I think was somehow caused by excess pressure in the tank since this completely went away after I removed the fuel cap gasket.

[Monkey Wrench]

modern cars frequently throw up warning lights to make the auto manufacturer money. If they wanted to, most of them could be eliminated with better use of modern technology.

[Chesterfieldway]

I would try copper sheet soldered to the tank--- if there is room enough.

[walfreyydo]

Just wanted to mention in this thread that I recently have resolved a "fuel smell" and "gas leaking from cap during autox" issue on my late model 944 S2.

The culprit was the late model plastic fuel filler neck. There is a metal sleeve which fits inside the plastic neck which the cap screws to. This metal sleeve had slightly separated from the plastic causing fuel that was sloshing inside the neck, mainly during autox, to leak past the gas cap seal. It was leaking out from between that metal sleeve and plastic neck.

Replacing with a early model one piece metal neck solved my issue. The conversion was pretty much plug and play but you will need an early model tank to neck collar, as the accordian style late model collar will block the one way check valve from opening and not allow you to fill your car with gas (actually, just very slowly).

The only way I was able to figure this out was remove the original fuel filler neck, attach the gas cap, fill with water and turn it upside down. To my surprise I discovered drops of water was able to leak past the gas cap where the metal sleeve and plastic neck meet. For a long time it mimic'd a gas cap seal issue, but after replacing both the seal and the cap, and still getting leaks, I began to suspect the fuel filler neck itself. My hunch was correct

[ESW83944]

Originally Posted by JonA

More to add on the over-pressurization of the fuel tanks: A month ago I removed the two way valve from the vapor line and drilled a small hole thru the plastic diaphram inside the valve. The thought was this would prevent the excessive pressure from building up in the tank due to thermal expansion/contraction. It seems to have worked, the tank does not build pressure/vacuum.
My concern was that this would cause a problem when filling the tank, that it would overfill. I have filled up 3 times since doing this and cannot see any difference, the pump fill nozzle seems to shut off just like it always did. The beauty thing is I can tighten the gas cap down now and all the fumes are gone.

Jon

Hi Jon, thank you for the great info! Looking to do the same on my 83. For reference, do you recall where you drilled a small whole in the valve? Thanks!

[Monkey Wrench]

I had to fix some rusted out gas cans, they were special cans that are worth a thousand or more.

I tried to clean the metal with a flap disk and braise them up from the outside. It worked somewhat but I probably could have done better paint removal. I did find that if the metal is clean and fluxed I could run the braise around and build the surface up fairly easily. the trick is not getting too hot and not burning through.

with mig welding yeas a guy can cut and replace sections but it's not going to flow and build thickness up without affecting the base metal like solder or braising rod can.

If I get mine out I thought I'd give it a good cleaning, maybe involve wet sandblasting or similar? dry ice blasting ? then use braising rod to strengthen any weak sections.

I got a kit to repair tanks Por 15 makes it. Id use it but not before completing any any welding or braising because if you coat it inside any heating will make it burn or flake off.

there are different fluxes for braising and some braising rod has the flux coating. silver solder is very similar. you may choose that.
I asked a pro welder and he said to find the "black flux" I went looking but could only find a white flux. he said the black flux has a wider heat range. if you overheat the flux it burns up but the metal needs to be basically near a red stage before it will melt the braising rod. Probably good to practise on some scrap tin first. It needs to be hot but not too hot so there is some practice and learnign curve with that. how the torch is adjusted can affect the flame temp. you can control the temp somewhat by tipping the flame away from the metal itermittently and watching how the braising rod melts.

plumbers sometimes use these plumbing torches, they run on acetylene not oxy-acetaline but I think that's hot enough for braising. I think the lower temp flame might be a little but easier to control and not reach the burn through temp.

if you try it using an oxy torch , its easy to get into a state where the thin metal gets so hot you blow a hole and then you have a larger issue. if the metal is clean and fluxed you dont overheat it then you can build the thickness considerably and strengthen the repair area with braising rod. one might find it a bit easier not to blow right through using a plumbers torch.

not all welders do a lot of braising, its used more for repairs than construction. I dont think it's really hard to learn to control the heat just takes a bit of practice. the por 15 seemed easy and worked. I think If I had my ( metal) tank out I'd do both the braising on thin rusted areas, the por 15 as a total coating after. what you do is put a little "extra in" then tip the tank to every angle imaginable to try to coat everywhere , then dump the excess out.

I thought about making a shaker by mounting a motor with an out of balance pulley on the tank and strapping the tank to a spring. my idea was to put small stones or small nuts in the tank and cause it to vibrate making the little rocks dance around and clean the tank from inside. It could be turned different ways and left for maybe an hour to do it's thing. the idea is just to clean it mechanically as well as chemically before any coating is applied.

one issue with sandblasting is it will thin the metal, it can also heat the metal and cause fractures by hardening it. using water with the sand , or dry ice blasting prevents overheating.

I think if you have a problem area, you can also choose to solder it and it's possible to take a thin section of copper and form it to fit the area, make it sit tight, then if it's clean and fluxed, then solder applied , capillary action and heat will draw the solder in between the brass or copper and the tank, that will also strengthen the area up. solder for this may involve lead solder and there may be a good tin to lead ratio, You can think about lead exposure if you are soldering with lead solder. Lead solder will melt at a far lower temp so you then wont need to worry much about bur through but you cna;t solder over rust, it needs to be cleaned really well somehow.

lead was once used for auto body work before bondo came into popularity and some body shop guys may still use it. one used to be able to find sticks of solder that were quite thick used for this purpose. I dont know if the trend continued but some restorers of fancy old cars took pride in doing any body repairs with lead, feeling it was less likely to fail than bondo. you may be hard pressed to find others with such experience today. - partly because of lead in the workplace causing health issues.


the por i5 kit comes with some cleaners, maybe if you are braising some type of autobody related acid may help remove rust from pitted areas prior to braising. trying to braise or solder or stick epoxy reliably over flakey rust will be a loosing battle. you cna convert he rust with a rust converter , that will hep somewhat if the rust is very thin.
If you do use the POR 15 I would not attempt any braising or welding after it's applied.

maybe its possible to send the tank out to have it hot tanked or dunked in something to remove the rust chemically, that might save work and exposure to nasty chemicals.

i was speaking to a guy that removed all the undercoat and rust and paint from his porsche, he used a sandblaster powered by a pressure washer. I thought it might be cool to try one. I have some old car fenders etc I'd like to clean and then work on.
an alternative is to use a bath and electrolysis to remove rust. Ive tried this uing a car battery or charger and some borax or washing soda.. It does work and eats away the rust with out pitting like acid, but i found it does leave a black deposit and I'd scrub it off between sessions in the tank. The problem it seems id that it would be difficult to mechanically clean inside a gas tank, a fender is different.

parts of a porsche might be galvanized, if you heat galvanized metal or put it in certain chemicals it can give off some really harmful fumes.

i threw a handful of fasteners that had a shiny cadium coating into some pickling paste. pickling paste is a very strong acid used for rust removal in commercial applications , its a very nasty acid. I was attempting to use the acid to remove debris from some parts I manufactured and braised and were blackened up, prior to painting and figured it might save a lot of fussy metal cleaning. I had to go to an industrial supplier to get it. probably best to not use that.

It got hot and started smoking off some vapors that were probably very harmful so it's worth being aware of that if you are messing about with these things. acids can trigger some chemical reactions and produce fumes that can be very hazardous.

I tried muriatic acid with similar results, its sold for cleaning concrete. HD sells it in gallons. it will dissolve the rust, it will also pit the metal, it will dissolve alunimum, it will remove all the paint, when you take parts out they are so clean that like with sandblasting they can flash rust almost immediately. I did clean lots of parts that way , I;d neutralize them and wash them and heat them with a torch and paint right away , it worked .. I'd dilute it about 1:3 with water.

It can work but It's stuff to be scared of. I tried cleaning a concrete floor with it and did about 6' square sections, after each section I used some baking soda to neutralize it.. I somehow managed to get a lung full of horrible fumes and oathed to never buy the stuff again.

[Monkey Wrench]

Originally Posted by ESW83944

Hi Jon, thank you for the great info! Looking to do the same on my 83. For reference, do you recall where you drilled a small whole in the valve? Thanks!

I think its intended that when you park your car in the sun and the tank gives off fume pressure that the fuel vapors condense and turn back to fuel in the charcoal cannister rather than simply off venting to atmosphere.

maybe when the tank cools that negative pressure sucks the raw fuel that condensed back to be burned by the engine somehow?

I think drilling out the valve disabled the charcoal canister basically. could this cause the charcoal cannister to fill with raw fuel after a while? ( not return ever maybe? )

I'm not sure. I think it is still venting to the charcoal canister since the cap has no vent, so far as I can reason. maybe what the valve does is causes the fuel to leave the canister when the tank is cooling and under a negative air pressure situation?

maybe we can understand better how it all works before we all go drilling those things out? If the cap wont seal that will also disable the canister as no fume pressure can build then so fumes wotn enter the cannister so much, they will leave via the cap.

could it be better to just disable the valve in the gas cap if the intent is just to vent it to atmosphere anyway? if the cannister fills with fuel over time, and has no mechanism to leave, that might not be very helpful but I'm not sure if that's the case here.