I've welded a bunch of fuel tanks- - - -even did a braze fix of a fuel fitting leak on top of one about half full of gas one day. The trick is to totally fill the interior of the tank with a non-flammable gas like nitrogen, CO2, or even exhaust gas from a very rich-running engine. No oxygen to support combustion- - - - -no kaboom. Pressure vessels don't have to have a fire to do damage- - - -the fracturing of the metal itself is the dangerous part.
Jerry
I haven't welded a fuel tank, but I supervised welding one from a gasoline engine M88 tank recovery vehicle (probably over 100 gal capacity). We filled it with water nearly to the level of the crack that needed welding, then discharged 2 CO2 extinguishers in it. When the welder touched the arc, it gave a little "boomp", at which point I told him to keep welding, as all the fuel had been consumed.
The problem with air tests on tanks or piping is that the air is compressible and thus stores energy. Water (hydrostatic test) on the other hand is incompressible, and the only energy stored is in the slight flexibility in the tank steel itself. Any rupture of the tank reduces the pressure nearly immediately, so no explosion.
Overheated water heaters, developing steam inside, however, can turn into rockets, and have destroyed housed in the process of exploding. Mythbusters did a video of one where they plugged the relief valve and let it cook. It took off and flew about 100 ft in the air thru two floors of the test house they had it in. Didn't find that one, but here's a video of one they did out in the open.
Water Heater Rocket Pretty exciting, but sobering too if you think about the damage it could cause.
