Repair Welding Percentage for Each Welder

By Tyrone

Date 11-05-2018 12:46

Hi Omid,
Are you asking how much each welder is allowed to perform a repair?

By khodabandeh (**)

Date 11-06-2018 06:33

Yes,exactly.As you know in most of the project it should be issued welding repair procedure document and one important things which must be specified is the welder repair percentage. So I need the suitable and logic way to calculating that along with relevant reference standard code.

By Lawrence (*****)

Date 11-05-2018 14:50

A number of the very large oil and gas players have a "typical" expectation of a 95% first pass yield for welds subject to volumetric testing (UT or RT).

Different players from that point have different solutions.

1) Retrain welder
2) Fire welder

Keep in mind the term "typical" Some projects have extreamly challenging working conditons and the expectation of weld test failures may be reasonably higher.

I'm not recommending you adopt the above. This is just some information from some of my peers.

By Tyrone

Date 11-06-2018 12:32

Okay, I see what he's asking now.
I doubt that any code or standard would specify allowable rejection rate for welders.

If I were you, I would gather statistical data on every welder (by weld length). Plot it out and see how close you are to 95% FPY. Then you have a decision to make on where to set the threshold.

Tyrone

By Shane Feder (****)

Date 11-06-2018 12:57

Omid,
Whichever way you choose to record repairs it is critical to evaluate the two scenarios first - both are used widely in all welding applications worldwide and there are always arguements over which is correct.
% of butt welds completed or % of total weld length completed.

I personally will always pick % of total weld length unless it is a pipeline where all diameters/thicknesses are the same for every weld.
Regards,
Shane

By khodabandeh (**)

Date 11-14-2018 05:47

Is it correct to calculate the percentage of repair with ratio of repair joints to total number of pipeline's joints?

By Tyrone

Date 11-14-2018 13:15

You could do it that way. It's all up to what you are trying to show.

(# of repair joint/total # of pipeline joints) will be higher than if you used: (length of weld repair/total length of weld joints).

I think the length of repair is a better method. More representative of the amount of defect.

By Shane Feder (****)

Date 11-16-2018 12:31

As I said previously it is always an issue whichever way you report it.
But remember this, and it is very important if your company takes quality seriously.
Don't just do things to keep the client happy - do things to ensure the repair rate of your welders (whichever way you report it) is acceptable to your company if you want to keep getting more work.
You need to ensure you have a team of welders whose repair rate is so low that it doesn't matter how it is reported - it will still be acceptable to the client.

My comment previously on pipelines was based on this - on a pipeline it is continual forward motion.
A repair - whether it is 25, 50, 100 mm long still requires someone to return to that weld and repair it at great expense.
So one repair in 10 joints is a 10% repair rate.

If you are in a refinery it is different - one welder welds 20 x 25 nb welds and gets 1 x repair (25 mm long).
Another welder welds 20 x 400 nb welds and gets 1 x repair (25 mm long).
Who do you think has the problem ? - based on welds they both have a 5% repair rate.
One has welded approx 1600 mm of weld, one has welded approx 25 mtrs of weld.

Your company is the one who sets the standard - if they can't meet your standard they either get retrained or fired.