Welding Repair Rates

Until either figure reaches zero, there is still a problem to solve.

So it doesn't really matter.

Do you work for the contractor, or does the contractor work for you?

Is the important number the number of failures/number of joints tested, or number of failures/number of joints welded?

Until you've tested them all it's not much use either way.

See PWSlack's comment.

OK, nice question, check the following points:

1.-It is important to categorize the fluids (Normal and M fluid, Severe Cicle, and fluid D). On it depends on the evaluation of the welds reject discontinuities.
2 .- It must be define the type of examination (Random, Spot , Random-Spot or 100%).
3 .- It must set the batch (Lot, see ASME B31.3 Parr.344 note 2 and 3) to be examined by welding, piping class and WPS (This shall be a number of welds (i.e 20 or 30 welds. Daily production or dates are not acceptable. See interpretations)
4 .- The percentage of rejection shall be recorded on these lots.
5 .- In case of random or spot, the progressive sampling of examination shall be according to ASME B31.3 Parr 341.3.4 . These additional shooted welds (Tracers) not count for the required percentage of examination but count for the percentage of rejection.
6 .- Applies to all required examinations. (RT,MT,PT and UT)

WP

Hi Whitephone,

Thanx for the comments. Gud and informative.

Regards,

Raj

Does no. of joints radio-graphed = no. joints welded ?

If the radio-graph testing is a spot check and not done on all the welds he is right in his argument.

Wrong in his to date assessment. Would only do it for some limit time period like on a monthly basis. Would show trend and whether they are getting better or worst.

Either way your the customer he is suppose to keep you happy. Apparently doing poor job in that regards.

Use it to your advantage. Tell him okay we will figure it your way but set a thresh hold you can be happy with and tell him if he goes over he eats the cost.

The answer to this depends on whether you accept the unchecked welds or not. Usually I would expect that a certain percentage of the number of fail divided by the number checked would trigger more inspection.

I have to ask who pays for the repair work.

In my experience, with structural steel, the more rigorous the inspection, the more the fabricator/erector charged. The harder you look the more blemishes you find, the more repair to pay for.

All of this should have been covered in the specifications and/or the purchasing documents. If it is not covered, you must negotiate.

I tend to agree with this statement:

"All of this should have been covered in the specifications and/or the purchasing documents. If it is not covered, you must negotiate."

All data can be manipulated to the advantage of the one analyzing it. Naturally one will argue for the most liberal interpretation to put him/her in a better light. That is the very reason that I agree with the statement above. If the specified method is agreed upon and followed, the subjectivity is minimized and the purchaser's intentions are satisfied. I further agree that if this information is not specified, you do then need to negotiate. The question then becomes a matter of whether one only wants to show his/her results more favorably or whether the prime concern is safety through quality verification.

I think the contractor's reasoning is wrong in this case. If you were to continue radiographing more (or all) joints, you would probably continue to find about the same rate of defects. (He is basically asking to let all the unchecked joints get by as freebies.)

contractor who claims that repair rate is "no. of joints rejected till date" / "no. of joints welded till date". If we calculate the repair rates by contractor's method, repair rates are very low, because many service lines have only 10%RT, 20% RT requirement.

Here you are confusing the matter. Do you consider 10%RT, 20% RT as not rephotographed? Please clarify. 10%RT or 20% RT is not decided by the contractor, it is as per contract and once it is completed you have to count as rephotographed. But if "no. of joints welded till date" includes the weld joints yet to be tested then contractor is wrong.

After all contractors job is to rectify all the rejected joints which is not going to change, whatever way you represent the repair rate. (why he should bother)

Mr rajehnair I don´t know if you have any agreemeent Customer/Manufacter, I only refer to the code.

Please review the following ASME B31.3 INTERPRETATIONS:

16-05

13-12

11-03

11-04

10-09

10-12.

Regards. WP