Repair or Replace after a Third Inspection

Date 07-28-2009 14:59

Good day Eric,

I am afraid that it's not something we simply choose, it is often written in either the code you are working to OR the clients spec.
Most codes I know "only" allow 2 repairs unless you have a specific repair procedure.

Seeing that you are welding P5 we need to know if the RT has been done before or after PWHT.

3.2

By rodriguezedb (*)

Date 07-28-2009 15:13

Thanks 3.2,

The RT was done before the PWHT.

Erik.

By 3.2 Inspector (***)

Date 07-28-2009 15:25

Erik,

Which code are you working to?

3.2

By rodriguezedb (*)

Date 07-28-2009 15:42

ASME B31.3.

Erik.

By 3.2 Inspector (***)

Date 07-28-2009 16:16

Erik,

I dont have 31.3 in front of me as I am on Bermuda right now :) (nothing beats the local rum)
But I would be very surprised if 31.3 allowed more than 2 repairs with the same procedure.

3.2

By paul prill (*)

Date 07-28-2009 17:43 Edited 07-28-2009 17:47

I have the 2004 B31.3 in front of me and 328.6 does not address the number of repairs that can be made, I would agree with the fact that in the contract documents it may address the number of repairs. Also the nature and extent of the defect(s) could have an impact weather or not this weld is a cut out or grind and reweld. the more you excavate and weld on a joint the harder it becomes to fix each time, especially if it was an open butt joint to begin with. Sometimes its more cost effective to just cut the darn thing out. In any event you may want to have a welder who is better skilled or experienced in repairs do the repair than the one(s) that made the original weld or attempted the repairs, some welders are just better at repairs than others.

By rodriguezedb (*)

Date 07-28-2009 18:40

Thanks a lot 3.2 and Paul,

Your comments have been very useful for me. I will look into the internal regulations of this company for this specific welding repair recommended practice. I hope to find it.

Erik.

By rodriguezedb (*)

Date 07-28-2009 18:41

3.2 Enjoy Bermuda!

By 3.2 Inspector (***)

Date 07-28-2009 18:44

I did, I am and I will in the future :)
I have been here 10 times now....

Glad I could be of just the slighest help to you.

3.2

By Joey

Date 07-29-2009 10:25

Hi 3.2

Now I believe that "You need a good amount of luck to get in the right places".

Any job vacancy there? Can I send you my CV? :)

Do they accept CAWI ? :) :)

Regards
Joey

By jon20013 (*****)

Date 07-29-2009 11:10

B31.3 does not specify a limit to the number of repairs however, given that P5 is considered a hardenable material, if it were on my job I would have it cut (replaced) including the heat affected zone and rewelded. Obviously you're doing the RT before PWHT as good practice and know that the "official" RT will need to be done after PWHT?

By 3.2 Inspector (***)

Date 07-29-2009 11:48

If it is not required by code, how would you get the joint replaced?

----> I find it a bit strange that the number of repairs is not mentioned in 31.3

3.2

By Shane Feder (****)

Date 07-29-2009 12:25

3.2,
I think the codes do not mandate a maximum number of repairs because there are so many different variables. Different materials can handle more heat cycles than others. Different types of repairs require different amounts of rework.
An example is an unacceptable gas pore in the capping run is going to require a lot less excavation and subsequent rewelding than a full root repair therefore less heat input into the joint.

Jon,
While I agree with what you are saying about the P5 being hardenable I think it would be dangerous territory to cut something out if there is nothing mentioned regarding rejection in the codes or the project specifications. Might even be a candidate for Geralds new favourite word - RAMFO. LOL !!
On a more serious note, a poorly written specification may make no mention of multiple repairs for P5 material but I would be very surprised if there was no mention of maximum hardness values. If the hardness is more than allowed then it is obviously a cut out whether it is the 1st, 2nd or 3rd repair.

A question for you Jon and the other gents with regards metallurgy, are there other reasons for not allowing multiple repairs other than increasing the hardness ?
Cheers,
Shane

By 3.2 Inspector (***)

Date 07-29-2009 12:48

Well....

More repairs does per default not make it more hard (actually the oppposite may be the case depending on the size/heat of the repair)
I would rather say that multiple repair makes it softer, again, it depends on the size/heat of the repair and ofcourse the cooling rates and other pre or post heat procedures.

High heat input = Slow cooling = softer steel.
Low heat input = quick cooling = Harder steel.

The above is only a rough guide as there are many things to consider in order to give a final answer.

Your example with the gas pore in the capping run and the little - or less as you call it - heat input actually has a big chance of hardening the steel as the cooling rate most likely will be rapid, another issue with small repair is the thermal stress you introduce to the weld during repair.

EN standars often describe smaller welds or tack welding requires higher preheat temperature than stated on the WPS, and the reason is the low heat input, and consequently higher cooling rate.

3.2

By Shane Feder (****)

Date 07-29-2009 12:55

3.2,
I may not have explained it well but all your responses validate what I was trying to say to your original question.
There are too many different variables to be able to mandate a maximum amount of repairs in the code.
Regards,
Shane

By 3.2 Inspector (***)

Date 07-29-2009 12:59

Shane,

I hear you....
That is why I think EN standards makes the right choce by saying "A Repair procedure is needed"

You can in theory repair as many times as you wish if you have qualified procedures in place, which demonstrates that the repair you are about to do has been tested.

3.2

By CWI555

Date 07-29-2009 13:22

It would in fact be trouble if you cut something out that is not required. The proper path is to perk it up to weld engineering with a clearly stated concern.
I would not want to sign off on that particular material without having raised the concern.

As for other reasons, they include enlarged grain structures, embrittlement (which is what the increased hardness is suggesting) and others.

Regards,
Gerald

By js55

Date 07-29-2009 12:28

jon makes I think the best point. It depends upon the alloy as to how many repairs you can allow.
The primary reason I'm sure the codes are reluctant to address it. A wise decision.
It also depends upon how the repairs progress. If each time the repair cavity is larger than before (not an unreasonable assumption considering the indication was missed the first time-unless of course it being put back in) then the HAZ and prior deposit will be removed, causing the repair to be as if it were the first one.
Sound engineerng judgement is indispensible. Funny how that keeps popping up.

By 3.2 Inspector (***)

Date 07-29-2009 11:46

J

It is not as good as it used to be, now everything is pre-fabricated in units.
10 years ago when I started everything was done on site.

We have done some good projects in various parts of the world, Barbados, Guam, Saipan, Sri Lanka, Mauritius, Bahamas, Cayman, etc...
Back then we did not pay tax :)

Now that things have changed, am I only on site a few days here and a few days there and paying almost full tax :(
So I am looking for something new myself...

3.2

By jon20013 (*****)

Date 07-29-2009 12:35

Shane, from a metallurgical standpoint, it all depends on the flavor of material. Numerous repairs can also damage HAZ after repeated attempts, in some cases (not the case with P5) the material may actually be softened to the point where tensile and yield properties are no longer what the were intended to be when in "original" state.

As for being dangerous to cut a weld rather than going into a fourth or more repair, of course this is a decision to be made by the engineer (or quality department in some instances). For me its simply a matter of practicality.

By Shane Feder (****)

Date 07-29-2009 12:48

Jon,
Thanks for the explanation, I hadn't even thought of softening the material.
My apologies on the stepping into dangerous territory, as the Welding Engineer you can cut out anything you want to- I just couldn't resist throwing in this new word - RAMFO. Now that has got me thinking - what does RAMBO actually stand for ????
Cheers,
Shane

By CWI555

Date 08-07-2009 15:34

Rambo - A surname chosen by a swedish immigrant to the United states.
Peter Gunnarsson chose the surname of Rambo from his place of origin, Hisingen, the northwestern section of Gothenburg, which is dominated by Ramberget mountain, meaning raven's or ram mountain depending on which dialect it was stated in.
He was actually an apple farmer, and introduced the 'rambo' apple into the colonial town of new sweden in 1637.

I believe it was the seeming duality of the original name that inspired the character name of the Rambo series movie.

However, the native people of the North American continent tie them together in legend.

"Bighorn Sheep (ram) were amongst the most admired animals of the Apsaalooka, or Crow, people, and what is today called the Bighorn Mountain Range was central to the Apsaalooka tribal lands. In the Bighorn Canyon National Recreation Area book, storyteller Old Coyote describes a legend related to the Bighorn Sheep. A man possessed by evil spirits attempts to kill his heir by pushing the young man over a cliff, but the victim is saved by getting caught in trees. Rescued by Bighorn sheep, the man takes the name of their leader, Big Metal. The other sheep grant him power, wisdom, sharp eyes, sure footedness, keen ears, great strength, and a strong heart. Big Metal returns to his people with the message that the Apsaalooka people will survive only so long as the river winding out of the mountains is known as the Bighorn River."

The Apsáalooke native americans are the source of that legend, their name was mistranslated by early interpreters as "people of [the] crows." It actually means "children of the large-beaked bird," a name given to them by their neighboring tribe, the Hidatsa. The Absaroka first encountered Europeans in 1743, two Frenchmen (the La Verendryes brothers from Canada), near the present-day town of Hardin, Montana. These explorers called the Apsáalooke beaux hommes, "handsome men." The Crow termed Europeans as baashchíile, "person with yellow eyes."

Between the Native American legend and the strangely similar legends surrounding rambo's (gunnarsons) home mountain of rambertget, it doesn't take much of an imagination to draw the inferences from both portrayed in the movie rambo character.

With that I would say 'Rambo' is not an acronym, but rather a history. Though I will admit it almost begs for an acronystic meaning.

Regards,
Gerald

By 3.2 Inspector (***)

Date 07-29-2009 12:51

jon,

That is indeed also the case with P5, P9, P12, P91, P92, X20, etc....

3.2

By bert lee (**)

Date 07-29-2009 13:20

3.2

what repair procedure must look for? do you have to prepare it or this is already inclusive in the quality mannual.
do you need to write a new wps? or the existing wps will do but with written statement on step-by-step method of repair work?

regards
bert

By 3.2 Inspector (***)

Date 07-29-2009 13:29

bert,

I LOVE THIS QUESTION!!

It's not written anywere in the EN codes how this procedure must be prepaired, as a consequence, nobody does it :)

When I am the client's rep. I always tell the contractor to PROVE to me the weld is OK.....I am yet to be convinced.
So they always cut it out - including the HAZ -----> which is IMO the best solution, but not required, which is why I ask them to prove it to me :)

Now the Bermuda Gosling rum is starting to make an impact on me, and the danish (company) internet modem has proberly exceeded $200 and I have a flight to NY tomorrow morning to catch....

Have a good one....

3.2

By bert lee (**)

Date 07-29-2009 13:45

okay, take care bro!

By js55

Date 07-29-2009 16:07

"I always tell the contractor to PROVE to me the weld is OK"

Considering that you can't PROVE that any weld is OK (barring a destructive test of course which would sorta defeat the purpose), this is a pretty safe approach.
My response would be "show me where in the specs it says that I have to prove a production weld is OK'.
Never seen it in no code neither.
If you take it to the engineer and he concurs, thats fine, then pay for it. Heck man, I'll cut out every weld on the project if'n the client wants to pay for it. :)

By js55

Date 07-29-2009 16:20

Also,
What is OK?
If I weld a Grade 91 with B3 filler using a Section IX qualified WPS, I can prove to you the weld is "OK" by tensiles and bends.
But is it?
If you make me "prove" its OK for creep service we'z gonna be waitin awhile.
And even then, depending upon the test temp and the lifetime requirement we may still be OK.
Having said this I think its insane to weld Grade 91 with B3 (except for dissimilars with Grade 22) but in many instances I can with a destructive PQR prove to you its OK.
Sound engineering judgment. Holy sacamolian!!! There it is again.

By jon20013 (*****)

Date 07-30-2009 03:13 Edited 07-30-2009 03:16

On my particular project of the moment, we are required by our specifications to have repair PQR's for mid-thickness and full thickness repairs. I have never done this in the past, but have to admit it's a good idea. And as js55 points out, almost all US Codes will tell the user right up front the codes cannot address every conceivable situation and yes.... sound engineering judgment is required :)

By jon20013 (*****)

Date 07-30-2009 03:21

Oddly enough, we have 2 piping specifications on current project, one for base code B31.3 and one for High Pressure B31.3. In the base code specification, it dictates that if there is a crack, the joint will be a complete cut out whereas for the high pressure specification, it doesn't mandate this. Obviously I have some heartburn with dictating a complete cut out if there is a "minor" crack on a large diameter, thick section. Unfortunately, this provision was adopted from a submersible pipeline spec and hasn't evolved yet.... next revision we'll put some common sense into it! :)

By js55

Date 07-30-2009 12:28

You do deal with some odd stuff jon.

By jon20013 (*****)

Date 07-30-2009 13:11

Never a dull moment js55! :) Headed home this evening, to Christchurch, leave my city here in Kazakhstan at 10:20PM and arrive home at 2:30PM Saturday afternoon.... its a long trek!

By js55

Date 07-30-2009 13:24

Its a lot of time in the air. Tell me you ain't doin this coach.
Man, that much time on a plane in coach I'd probably never be able to walk again.

By jon20013 (*****)

Date 07-30-2009 14:59

Cattle class dude. If I'm lucky, I'll get an aisle! :) Last time back to the USA, I got really lucky, upgraded to Business Class both ways across the pond. No chance it'll happen two trips in a row! :(

By js55

Date 07-30-2009 16:25

You're tougher'n me.

By CWI555

Date 07-31-2009 02:17

cattle class? LOL

By 3.2 Inspector (***)

Date 08-02-2009 17:01

Which is why client's who deal with P91/X20 on a regular basis has specified which type of filler material is allowed (no we dont have 100.000 hours to wait)
I can imagine P91 is not being used except if creep is an issue, so whats your point?

3.2

By js55

Date 08-03-2009 12:36

Really? You don't get the point?

By 3.2 Inspector (***)

Date 08-03-2009 12:38

Funny you say that, I was about to write the same to you :)

3.2

By 3.2 Inspector (***)

Date 08-02-2009 16:57

Why would I pay you to cut out welds?

If you repair it more than allowed by code, yes you need to prove it still is OK ---> or cut it out.

"My response would be "show me where in the specs it says that I have to prove a production weld is OK" And my simple answer to that: Show me where it is written that you are allowed to make this 3rd (or more) repairs.

Since you do not have a qualified procedure, how can the weld be valid?

3.2

By js55

Date 08-03-2009 12:39

"Show me where it is written that you are allowed to make this 3rd (or more) repairs."

Are you saying that if the code doesn't explicitly address something and approve of it it is prohibited?
If so, where do the statement of sound engineering judgement come in. If the code is a cook book who needs engineering?

By 3.2 Inspector (***)

Date 08-03-2009 12:56

NO!

Early in this thread I made ref. to EN standards
In the EN standards it is clearly written that only 2 repairs can be done unless you have a special/modified repair procedure.

Most contractors do not have this special/modified repair procedure ----> This is when I ask to show me - or prove as I wrote - that the weld is OK.
Maybe "OK" was not the right word to use.... They have to prove to me that the weld, which they repaired 3 times (without correct procedures) is still valid, which they can't.

Remember when I asked you about some P91 structure some months ago?
Futher investigation (destructive) showed that the weld no longer had the properties needed for super heated steam for the intended life of 15 years.

That was only discovered because I was SO stubborn that I requested - for 3 months - the contractor to prove to me that the weld was OK (or whatever word you want to use here) They had all kind of people telling me that everything was OK.....but I wanted proof, which they could not give me. Finally the client ordered replica and later destructive tests.

"Are you saying that if the code doesn't explicitly address something and approve of it it is prohibited?"

No, but when the code CLEARLY prohibit something, they better show me that what they did is acceptable.

3.2

By js55

Date 08-03-2009 13:03

"Why would I pay you to cut out welds?"

If you tell me to cut out welds that are acceptable per code, per my approved WPS, per my manual, and per the contract, you're paying for it.

By 3.2 Inspector (***)

Date 08-03-2009 13:06

I think I asked you this before: WHY would I tell you to cut out welds which are done according to code?
Please read the part where I make ref. to EN codes.

3.2

By js55

Date 08-03-2009 14:30

Ah. OK.
I now see where the EN codes constitute "most codes you know". :)
My misunderstanding.

By 3.2 Inspector (***)

Date 08-03-2009 14:32

"That is why I think EN standards makes the right choce by saying "A Repair procedure is needed"

No problem at all....

3.2

By js55

Date 08-03-2009 14:32

If they are done according to code then why would I need to prove to you they are OK?

By 3.2 Inspector (***)

Date 08-03-2009 14:34

I refuse to answer that.
My memory will also be worse as I get older.

3.2

By rodriguezedb (*)

Date 08-06-2009 15:47

Thank you all (very much) for sharing your experiences with me, was very helpful. Here are some facts and decisions taken in this case:

- The joint (weldment) was replaced, the filler metal and HAZ was removed.
- We used the same-WPS for the completion of work.
- Prior to the cut, we made a hardness test and in situ metallography. Was reached that the microstructure of the HAZ and base metal in the area that had been repaired three times had some irregularities (grain growth, low hardness).
- In line with the codes, there are no internal regulations (in the company I work for) governing the amount of repairs. Currently I am working on it.
- The contract stipulated that the contractor will be responsible for all costs that are generated after a second repair, ie, the owner recognizes only two repairs (I didn't found the technical bases of this clause).

Thanks again for your time and comments.
Erik.