Extent of required examination. B31.3

Date 03-17-2005 04:58

I’m looking at B31.3 (1996) paragraph 341.4 concerning the 5%. If you RT 5% and have some failures do you still need to RT some more welds (other than the repaired welds)? My first place to look was at the contract requirements and our other facility specifications but the only thing mentioned is 5% of fabrication. I’m not sure if that is by design or error.
Regards, Donnie Mann

By thirdeye (***)

Date 03-17-2005 07:49

Donnie,

Paragraph 341.3.4 (2002), describes the progressive sampling requirements when a random examination reveals a defect. As you can see, this penalty pyramid can quickly get out of hand. The welders refer to it as "three strikes and you're out".

~thirdeye~

By dlmann

Date 03-17-2005 08:34

thirdeye: Thanks. There it is!!! Right on the same page as I was reading but in a different paragraph. The hazards of reading code books while on night shift rotation!!!! What is your interpretation between 341.3.4(e)(1) and (2). I don't know the difference between " repair as necessary" and "repair as required".
Regards, Donnie Mann

By thirdeye (***)

Date 03-17-2005 17:13

Donnie,

In 341.3.4 (e)(1) “Repaired…. as required” is referring to the entire lot (minus any acceptable welds in the original 7). That is the penalty for a failure in the third tier (the base) of the pyramid.

In 341.3.4 (e)(2) “…as necessary” is used due to the fact that the lot has undergone a full RT examination. It would only be necessary to repair or replace any defective welds.

~thirdeye~

By NDTIII

Date 03-19-2005 07:49

You also need to check your engineering design package to see if they require additional examinations.
Normally the engineering design, the contract or maybe even the purchase order will require more than 5% and will delineate the percentage as well as the percentage of what? Line class? System? Isometeric? It depends.
Engineeering design may require 100% RT on Hydrogen Service for instance.
Don't rely solely on that statement in the code.
Thirdeye is absolutely correct on the progressive sampling requirements. It can get out of hand very qiuickly.