We want to consult you about a recurrent welding failure problem occurring at a hot water line. The water is heated in a furnace to 330 °F (165 °C) and pressurized to 350 psi, then is sent through carbon steel pipes to closed heat exchangers where it is used to increase other fluids temperature through closed heat exchangers, cooling to 258 °F (126 °C) and returning to the furnace. The water circulating through the line is demineralized and runs in a closed cycle, so supposedly it doesn't suffer contamination. However, the water chemical analyses have showed high hardness, alkalinity and pH at certain dates and the analyses also have showed residual concentration of Eliminox (NOx and CO combustion promoter for fluid catalytic cracking (FCC) applications). The line was built of carbon steel: seamless A-53 Gr. B pipes (2", 6", 8" and 12" diameters), seamless A-234 WPB elbows and A-150 flanges. All the line has a 4" thickness thermal insulation.
There have been eleven welding failures in two months. Ten of them occurred at joints welded using E-8010G for the root pass and E-7018 for filling and cap passes. Ten failures occurred on pipe-elbow joints and one on an elbow-flange joint. Ten failures occurred at the furnace exit at 330 °F(126 °F). Seven failures were longitudinal cracks at the welding, one failure was a transversal crack and the remaining three are leaks at pores. Of the failures five occurred at 12" joints, three at 8", two at 2" and one at 6". The welding joints were done in-shop and at the field by different welders.
Thank you for your help in solving this problem.
Cordially,
Monika Gomez-Duran
what is the reason for using the 8010 for the root
my bet is the 8010 root. I would bet that the welders didn't preheat the joint before welding because of the 7018 fill and cap. and the root cracked away from the base metal.
You are overwelding the pipe!. You need to use 5p UH for the stringer and 7018 out. The 5p is very ductile where the 8018 is not. If they are running the bead DH, then 45 minutes or longer passes, when you hit it with the heat "SNAP". You did qualify a procedure on this did you not? There are several things, such as the preheat to 200 degrees, maintain interpass temp, time between bead and hot pass that need to be checked.
By MBSims 
Date 06-22-2007 03:38
Edited 06-22-2007 03:58
Has the line been evaluated by an engineer for thermal expansion stresses? I say this because you said there were 10 failures at elbow welds and 1 at a flange joint. The elbows will be high stress locations. There must be sufficient flexibility in the piping system design to accomodate thermal growth without exceeding the allowable bending stress. Have a qualified piping engineer do a flexibility analysis and evaluate the supports.
You can find some basic info at the following link. Be sure to follow the instruction to scroll down, click on free, enter the download code shown.
http://rapidshare.com/files/37140995/Pipe_Stress_Course.zip.html
Why would there be Eliminox inside the water piping? Are there a leaks inside the furnace (boiler)?
By CWI555 
Date 06-23-2007 16:18
Edited 06-23-2007 16:52
I have to agree that it is overwelded, but i think you are on to something. There's more to it, there should not be any eliminox in the line.
edit: SA 53 grade B has a tensile of 60k, 8018 is a bit harsh. I'd be surprised if there were no leaks given than.
