welding over a cracked weld... sounds wrong

[OneDay11]

Hey guys!

A friend of mine (yea right) was talking about a cracked weld that was found on an offshore platform flair boom during a shut down. I don’t know the details but this is one of the structural welds and I am assuming it is one of the welds that connects one of the diagonal tube members to a horizontal tube member on the boom truss. I am not sure if the welds are full penetration welds but I am assuming they are. Anyhow, a 40mm crack was found and in the end they decided to just weld over the crack as a repair. Something about this just sounds really sketchy. Should they not have gouged out the old weld material and just done a brand new weld?? It will be a while before they do another shut down so a new repair will not happen for at least a year. Would love to get some feed back.

 

[SJones]

Was your 'friend' the welder, the inspection engineer, or the structural engineer directly involved with this repair? Or, was he the receiver of the story second or third hand? It may not have been necessary to remove the complete weld depending on the nature, depth and location of the 'crack'. Certainly, some form of repreparation should have been performed so as to be confident that the defect has been completely removed and not simply welded over to reoccur again. Welding Handbook, Vol 3, Chapter 9 Maintenance & Repair Welding gives a good overview of a model process that could have been followed. One hopes that the reason for the crack has been identified and steps taken to guarantee future structural integrity.

Steve Jones
Materials & Corrosion Engineer

http://www.linkedin.com/pub/8/83b/b04

 

[Metaljon]

It is not normal practice to weld over a crack but sometimes it is not possible to fully remove a crack and a temporary weld repair is performed knowing that some part of the defect is still present. This is usually backed up by defect assessment that ensures that structural integrity is maintained until a more permanent repair is planned in the future.

http://nz.linkedin.com/pub/jonathan-smith/39/25/589

 

[OneDay11]

This came from a guy with a background in materials and inspection. Most likely 2nd hand info. So yea... I take it with a grain of salt. I am just getting ready for the day that I would face a similar situation directly. I will check out your welding handbook reference. I also hope that proper measure were taken
Thanks for the insight!

 

[Duwe6]

On the Nuclear Powerplant side of things, there actually is a researched and approved "Embeded Flaw" technique that allows you to leave most of the crack behind.

On thick vessels, the cracked surface is excavated about 3/8" down and a total width of 3/8" [3/16" either side of the crack, and past the ends of the crack about 3/8". The mechanism is that when this partial-penn weld is made [in a specific sequence, with specific heat inputs [bead widths], the weld shrinkage will pull the edges of the crack togather. Saves a deep excavation that would cause PWHT to have to be used.

Problem is that the crack has to be FULLY mapped using a highly skilled UT tech, and a lot of Finite Element analysis done of the stresses in the region of the crack, and a fair ammount of specific, weld bead-by-bead Weld Engineering. Then everly 1-2 years [every opportunity] that crack has to be fully resurveyed, using that highly qualified UT tech. This is very Engineering intensive $$$, and adds one more mandatory inspection area for every shut-down $$.

The other Embedded Flaw technique is simpler, and only applies to socket-welds. The weld is peened shut, while in operation, using a center-punch until the [water] flow stops. A root of E-6010 is woven over the peened crack. Then the same ammount of weldmetal is applied to that weld as was initially required. Thus if the socket-weld size was 1/4-inch, there will be the original 1/4-inch weld, the 6010 'root', and another *full* 1/4-inch of weld layered onto the socket fitting and pipe. It makes a horrible looking knot of weldmetal, but it is EPRI approved and seems to work well.

Do not attempt the second one with onstream-flammable [or combustible] fluids.

 

[OneDay11]

@Duwe6
quite intersting stuff! do you have any references that explains the processes further?

 

[Duwe6]

Not personally, sorry. Both processes were developed by and tested at the EPRI.

Me, I'm a "gouge the crack ALL the way out" guy. I would use the "peen it shut" technique for a *temporary* repair, to limp along until the plant can be shut down.

 

[Guest102023]

I am of the same mind as Duwe6: do it for a temporary fix only, and not without some engineering justification. For CYA if nothing else.

 

[Guest102023]

CYA, or CYC ... Cover Your Crack, so to speak

 

[Duwe6]