How header pipe surface crack, after 5 years later on the Weldolet Weld Finish Area

[Selobay]

We have a problem on WELDOLET connection area on the header pipe. After 5 years later, here is a leakage and when we investigated it, we saw the surface of header pipe was cracked. The cracking location is just the end on weldolet "weld bead reinforcement area".

While I searching the reason, I saw the same problem on KevinNZ 's photo.

What is your opinion, why here is a crack ?

*INFORMATION OF PIPE & FITTINGS LINE*
Header Pipe Size: 20" SchXS
Weldolet Size: 3" 3000 LBS
Material: Carbon Steel
Fluid: BUTANE , 135°C @ 24bar

thread378-446206

 

[Snickster]

I believe it may be due to fatigue failure at high stress concentration point of the weld area. To avoid this weld contours are made curved and smooth with the base metal parts.

 

[SJones]

What was hanging off the 3-inch branch? Valves? How many? Has there been an Energy Institute VIF assessment?

Steve Jones
Corrosion Management Consultant

http://www.linkedin.com/in/drstevejones

All answers are personal opinions only and are in no way connected with any employer.

 

[GD2]

OP,
I agree with Snickster. High probability of fatigue failure. Process vibration induced fatigue failure in Small Bore Connections (SBC).
Try provide bracing/ gussets after the welding.

GDD
Canada

 

[mfgenggear]

Op
Has this been weld repaired?
I am sure corrosion stress added issues as well

 

[Selobay]

What was hanging off the 3-inch branch?​

There is only one valve. That 3" branch is for a BY-PASS line.

Has this been weld repaired?​

No there is no repaired. That weldolet was just removed from their weld and seen that crack area on top of the pipe. There is no any crack at the weld area. Seen crack was occur on the surface of pipe.

 

[r6155]

1) This is a sockolet (not weldolet). This is not a good design in this case- The lack of gap in the socket fillet weld may be a cause of this failure.
2) Excessive weld with the header and not a nice weld. Need a radius to a smoth finish
3) Do you have a complete report during fabrication and inspection?.

Regards

 

[Selobay]

Thank you for your correction.
As I understand your commend is, those Excessive Weld and weld quality not good.
They couldn't give me an inspection report unfortunately : (

 

[Heaviside1925]

OP,
I agree with r6155. Addition comments though it's difficult to be 100% based on the pictures.
- It appears the hole in the run was made the same OD as the o-let, based on the 1st and second picture and the grinder marks on the second picture.
- I am not seeing indication of a root gap nor full pen weld
- The overlapping 4 weld beads I am seeing should not the pressure retaining weld and as r6155 noted, should have been fillet with a smooth finish.
- There should have been a weld map or branch welding standard the welder should have been referencing, so not only the welder did a poor job but whoever was checking their work failed as well.
- If this was just a one-off maintenance weld, I'd have it repaired properly and move on, but if this was performed as part of a larger project, I would be very suspect of other welds in this installation.

 

[NovaStark]

@OP

Do you have any other information about the line or has your investigation determined it to be due to the welding ?

 

[r6155]

I suggest the nozzle NPS 3 inserted through the pipe wall, no sockolet / weldolet., for this example.
Regards

 

[mfgenggear]

R6155
All I can determine is that the wrong weld procedure was used. Why no olet.
According to My research olets are common and simple to cut prep and fit, then weld.

Major cause of cracks is to much heat. And
To add the rapid change in temperature.
Years back we would preheat with rose bud torch. But jump back when it was critical for the weld prep to be free of oxides. Contamination. Burrs, eg. And the gap was set.
I see welders welding with out a WPS. And asking if the weld was done properly.

 

[KevinNZ]

IMO O-lets welds can put too heat into the header. O-lets are designed assuming the branch and header are the same sch. e.g DN100 Sch 80 (8.56mm) to DN500 SCH80 (26.19mm). Now that is a lot weld if the header is STD(9.52mm)

 

[r6155]

Remember: this happens 5 years later

(OP) says: Material, carbon steel
This information is vague, I need the exact material used to continue.
This is a brittle fracture and resistance to the impact test is critical.
If there is no further information, then a laboratory analysis is required for each component, including the weld material.

Regards