A weld on a 2018 power plant job made me rethink our whole pre-heat process

We were on a big outage in Georgia, working on a superheater header. The spec called for a 250 degree pre-heat, which we did. But the first pass on a thick section still cracked. The old foreman on site pulled out his temp stick and showed me the base metal was only at 180 in the center of the bevel. We had to grind it out and start over, adding a full hour of soak time. Now I argue we should always verify internal temp, not just surface. But some guys say it slows the job down too much for a maybe. Has anyone else had a weld fail from a bad pre-heat reading?

3 comments

3 Comments

spencer_perry7710d ago

Yeah those temp sticks are good but they only show surface temp too, same as a gun. The real trick is drilling a small hole in the bevel face and checking the temp at the bottom of that hole, that gives you the internal temp where it actually matters. We started doing that after a similar crack on a thick wall job and haven't had the problem since. It adds maybe 10 minutes per joint but beats grinding out a full weld.

the_julia1mo ago

Forget slowing the job down, what's the real cost of a failed weld? That hour of soak time is nothing compared to rework or a shut down later. How do those guys measure pre heat now, just a surface probe?

shane2441mo ago

Remember seeing a crew try to skip the soak on some thick wall pipe. They just waved a temp gun at the surface like @the_julia said. It passed their check, but the weld cracked straight through after a week. The shut down to cut it out and redo it took three days and cost more than the whole original job. Now they use those stick on temp tabs that show the actual metal temp all the way through. It's slower up front but man does it save pain later.