Welding a fuel tank

[GNM109]

If there is no air/oxygen there is no flame/explosion. IF the motorcycle gas tank was full of water with the place to be welded submerged where was the air/oxygen for the explosion/bang you mentioned that sent water/fire 30 feet into the air?

What about the method of welding on a gas tank full of gas? If explosions happen while submerged why are the guys that weld on full tanks still around to talk about it?

You said you were a younger kid at the time and now retired, so it must have been some many years ago. Possibly you missed when he dumped the water out of the tank before welding it?

No, the water was not dumped. Flame exited the fuel filler opening.

Anyway, without regard to what happened in my report, it's nice to know that there is no longer any danger in welding fuel tanks. No need even to clean them. Just fill with water and have at it.

There's nothing more to say here. You have solved the matter. Carry on.

[vpt]

No, the water was not dumped. Flame exited the fuel filler opening.

Anyway, without regard to what happened in my report, it's nice to know that there is no longer any danger in welding fuel tanks. No need even to clean them. Just fill with water and have at it.

There's nothing more to say here. You have solved the matter. Carry on.

Thats what I do, but I am not liable for anyone else that does it.

Strange the filter housing wasn't full of water being the tank was full unless it was after a valve in which case the flame/explosion shouldn't have entered the tank sending water 30 feet high. We must get to the bottom of what happened here.

[Magicniner]

Some say that boiling/steaming out tanks doesn't get all the gas out of the seams.

How do you check for leaks in a tank after welding? Then if there is a hole what do you do?

Intelligent testers will use the lowest viscosity fluid cheaply available (air) to pressurize the tank to a few psi and use a leak test dip or fluid, easily finding leaks so small the water or fuel might not be immediately visible, it's not rocket science,

;-)

- Nick

[Magicniner]

Exhaust gasses from a gas engine are not always inert.

But they are low enough in oxygen to guarantee that an explosion will not occur and using piped-in exhaust provides the vital turnover of gas content to ensure the gas in a container stays non-combustible.
Plus it saves you stupidly putting water in a vessel with seams which are hard to dry out and vulnerable to rusting, results can be quickly tested gas-tight and any leaks fixed and the tank doesn't need drying out once it tests good,

;-)

- Nick

[vpt]

But they are low enough in oxygen to guarantee that an explosion will not occur and using piped-in exhaust provides the vital turnover of gas content to ensure the gas in a container stays non-combustible.
Plus it saves you stupidly putting water in a vessel with seams which are hard to dry out and vulnerable to rusting, results can be quickly tested gas-tight and any leaks fixed and the tank doesn't need drying out once it tests good,

;-)

- Nick

Till a guy with a 1962 chevy with no cat and a bad tune on the carb fills a tank with a rich gasoline mix.

[J Tiers]

If it was "full of water", then where was the air and fuel to MAKE the explosion?

OK, you were there.... So what?

Obviously some part of the story is not true. That does not make you a liar, it makes you a person who didn't have all the information.

The tank obviously was not "full of water" as you thought it was. There had to be some air and fuel mix in it, or no explosion was possible. It did explode, so it WAS there, and the tank was NOT "full" of water. Done.

************************************************** *************

Just for the sake of curiosity..... I did a little figuring

Assume a gas tank in the form of a 1 foot cube.... with residual gasoline a thou thick over the surface. That's not a lot, but it should still be visible.

Gasoline has a stoichiometric ratio of 14.7 to 1 in mass. Limits are about 1.5% to 7.5% for explosive mixes, by volume of gas (harder to figure than masses/weights) The tank holds air of about 0.08 lb weight. A uniform coating of gasoline all over the inside of the tank of thickness 0.001 inch would have a volume of 0.001 x 144 x 6, or about 0.86 cubic inch. The fuel weighs .026 lb per cubic inch, so there is about .022 lb of fuel involved.

Ratio of air to fuel is .08/.022 or about 3.5 to 1. Way richer than stoichiometric if it all evaporated into the volume of the tank. Obviously, depending on how much of the film of fuel evaporates, you can get a variety of ratios that will span the flammability range up to what is likely to be too rich. So it does not take much fuel in the tank to set up an explosion.

Even in the case of insufficient fuel to mix with the air and be in limits, the fuel may not necessarily mix evenly, so a portion of the volume may be explosive, and the rest too lean. Still may be enough to pop the tank.

[Peter.]

I have cut the bottom out of three aluminium motorcycle tanks and welded in an extension to hold more fuel. Each time I removed the filler, removed the fuel pump (on a 6" diameter plate which un-bolts from the underside) then I left them out in the sun for a couple of days. Didn't cause me any dramas or so much as a gnat's fart of a bang.

I don't think I'd be so care-free with a typical car fuel tank with the long filler neck though. I think they would need more work to get the combustibles out.

[Dieseldoctor]

If you talk to the family and friends of anyone killed or hurt by a fuel tank being welded on you will always hear, "he has done that for years" or "he has welded hundreds of those with no problem". Very few people get hurt at their first rodeo or first car race etc. Just saying.

Now I had a friend years ago whose job was to keep chain saws, axes, etc. in operating shape for a crew of about 20 woodcutters. When his metal gas cans would get all dented and even crushed from throwing around he would line them up, take all the caps off, and then take a torch and go around sticking it in the neck of the can. Can would jump and a flame about a foot long would fire out both openings but the dents etc. were usually pretty well ironed out. Never saw a can fire off but one time which evidently took care of all the fumes.

I have only worked on lawnmower gas tanks and very few of those. I either hook a hose to the outlet (with cap off of course) and let approximately 5 PSI of air circulate through the tank as I solder (with a big electric soldering iron) or if I can get my hands on some dry ice I put several chunks of it in the tank and just a little water to help it melt. When the "fog" starts boiling out I know it is safe to have flame around the tank. But I also do like the man I mentioned above. I stick a flame from a torch into the filler to see if anything with light off whether I am using air or dry ice. Never even had even a little pop but that's not saying the next one won't take my head off. But I still only use the soldering iron to actually do the work with.

Now the last one I did was my own tank off a Cub Cadet and I used JB Weld and it has not given me any problem after about two years. Sanded around the hole real good. Used thin layers with drying time between layers. Your outcome could vary.

[Iceman Motorsports]

Guy with a machine & fab shop in the area was welding on a diesel tank, as he said he has done hundreds. In the end he got a short hospital stay and he had great insurance, got a new shop and all new equipment, much better than he had before. He told me the guy swore it only ever had diesel in it but doubted that was true.

[Willy]

He told me the guy swore it only ever had diesel in it but doubted that was true

It probably was true and only did ever have diesel in it.
Anybody that thinks they are safe welding a tank because it has "only" had diesel in it probably should not be welding tanks. It's a motor fuel, why would he not think that it couldn't form a combustible mixture?
There are literally hundreds of compounds, liquid and solid that given the right conditions are capable of producing a lethal explosion.

[GNM109]

If it was "full of water", then where was the air and fuel to MAKE the explosion?

OK, you were there.... So what?

Obviously some part of the story is not true. That does not make you a liar, it makes you a person who didn't have all the information.

The tank obviously was not "full of water" as you thought it was. There had to be some air and fuel mix in it, or no explosion was possible. It did explode, so it WAS there, and the tank was NOT "full" of water. Done.

************************************************** *************

Just for the sake of curiosity..... I did a little figuring

Assume a gas tank in the form of a 1 foot cube.... with residual gasoline a thou thick over the surface. That's not a lot, but it should still be visible.

Gasoline has a stoichiometric ratio of 14.7 to 1 in mass. Limits are about 1.5% to 7.5% for explosive mixes, by volume of gas (harder to figure than masses/weights) The tank holds air of about 0.08 lb weight. A uniform coating of gasoline all over the inside of the tank of thickness 0.001 inch would have a volume of 0.001 x 144 x 6, or about 0.86 cubic inch. The fuel weighs .026 lb per cubic inch, so there is about .022 lb of fuel involved.

Ratio of air to fuel is .08/.022 or about 3.5 to 1. Way richer than stoichiometric if it all evaporated into the volume of the tank. Obviously, depending on how much of the film of fuel evaporates, you can get a variety of ratios that will span the flammability range up to what is likely to be too rich. So it does not take much fuel in the tank to set up an explosion.

Even in the case of insufficient fuel to mix with the air and be in limits, the fuel may not necessarily mix evenly, so a portion of the volume may be explosive, and the rest too lean. Still may be enough to pop the tank.

There was obviously some petrochemical residue that the torch ignited, possibly in the crack at the lower right hand corner of the tank. The tank was indeed full of water and had been exposed to a car exhaust for a few minutes before the welder filled it with water and got started. There was an explosion, actually two, although there was no damage and they exited through the gas tank filler hole. Obviously I was there and you were not but thank you for not calling me a liar. Your post would have been more complete had you done that, however since others on this site don't often hold back as I've seen in many other threads here having nothing to do with myself.
.
The remainder of your post addressing issues of fuel to air mixtures, although interesting is irrelevant in this context.

Done.

[J Tiers]

I suppose the point is that there cannot be any explosion unless there is air, fuel, and an ignition source. If full of water, there can be fuel and ignition, but no air. Water is not air, so there could be no burning.

Presumably everyone thought it was full of water. Myself, I suspect there was an air pocket at the weld location, that supplied the oxygen for the explosion. And there could not have been much else between the pocket and the filler neck, because otherwise there would have likely been a seam split, or at least some distortion of the tank from internal pressure.

Welding otherwise than backed by air would have developed a bubble of steam there. You can weld underwater, even stick weld, I am told. I know cutting torches can work underwater. But you can't get things to explode in a steam atmosphere, the oxygen is already bound to hydrogen, forming its oxide (water).

[GNM109]

I suppose the point is that there cannot be any explosion unless there is air, fuel, and an ignition source. If full of water, there can be fuel and ignition, but no air. Water is not air, so there could be no burning.

Presumably everyone thought it was full of water. Myself, I suspect there was an air pocket at the weld location, that supplied the oxygen for the explosion. And there could not have been much else between the pocket and the filler neck, because otherwise there would have likely been a seam split, or at least some distortion of the tank from internal pressure.

Welding otherwise than backed by air would have developed a bubble of steam there. You can weld underwater, even stick weld, I am told. I know cutting torches can work underwater. But you can't get things to explode in a steam atmosphere, the oxygen is already bound to hydrogen, forming its oxide (water).

It's possible that there was an air pocket. We will never know. I can only tell what happened.

[flylo]

No gas for 5 years I'll weld it. I've welded many by dumping out the gas & running a hose from an exhaust pipe purging all the air out of the tank. In the rental business the insurance company insisted when a machine came back to fill the tank before putting them back on the showroom floor. They were smart enough to know that a full tank is safe & a almost empty is not.

[Don Young]

In the 40's and 50's, Harley and Indian tanks were small and strong enough to contain any fire/explosion. It was common practice to empty the tank, blow it out a bit with air, and stick a lit torch in the filler opening. The result was a pretty good boom but I never saw any evidence of tank deformation. When the torch was applied a second time there was often a second smaller reaction. I have no idea how close the tanks were to rupturing and certainly do not claim this procedure to be safe.

I have a 300W American Beauty electric soldering iron which I have used to solder automobile gas tanks. I won't get a torch close to one.