We use all the typical CC/CV setups in under water welding, red, blue, yellow, grid power, gas/diesel, etc. The DC isn't a problem, you will probably get "fuzzed" a few times doing the welding, not buzzed...fuzzed. Underwater it is different, we do not hook a ground up to the welded component, just attach the ground to a plate and drop it about 10' down in the water column. The only thing that goes down to the work site other than the umbilical is the stinger. There is no real issue, every time the lead goes "hot" (knife switch is closed to provide current) I would get a fuzzy tingly feeling. Any cuts, scrapes, or anything like that would make things interesting, the old filling in my tooth would sure let me know it is still there. Overall nothing too harsh, unless you get in between the path of electricity from the rod back to the grounding plate, that will wake you up. DC is fine in wet conditions, DCSP or DCRP, typically I have always used DCSP wet. Never AC welding in wet, obviously (I hope). Anyway, the rods I used for practice and small welding stuff was just 7018 heavy dipped in wax, the flux heats up and the rod burns fine underwater, the flux does the same thing underwater as it does in air. Normal wet rods are different, but similar to 7018 I think. One thing to note, the Hydrogen effect is still there from the moisture. Wet welding is an entirely different code than the standard AWS D1.1, wet welding falls under AWS D3.6. Which reduces the strength of the welded component.
There are several variations of structural beam that are suitable. H beams in 80' span will have to be heavier, I beams "S" shape will be similar. The typical web of an H beam can be reinforced on both sides by plate steel to span longer distances while supporting the same load, a cap channel can be added to add additional strength to span, or flange-to-flange stiffener plates can be welded in to support the member. The beam doesn't need to be heavier, but does need to be taller. Any engineer that is worth a damn would know all of this, the AISC manual spells it out pretty well. A fabricated truss is not a bad idea, there is just a lot more engineering that needs to be done to account for the unknown properties, unlike mill steel which is well known and documented properties. Good luck