How close by code can two welds be made to each other?

[bizsheets]

I have a customer questioning the distance between 2 butt welds on a piping system (photo available) designed (customer spec'd) to B31.1. While they cannot cite a code that prohibits it, they state: "The two closely-spaced weldings is NOT a good practice. Generally there should be enough gap (space) so that Heat Treated Regions do not overlap each other." and are referring to ASME Boiler & Pressure vessel code , vol. VIII Part UW REQUIREMENTS FOR PRESSURE VESSELS FABRICATED BY WELDING Serie UW-9 , Design of welded joints , paragraph (d).

Question is:
• What, if any, is the relevance to the ASME boiler code VIII section listed above to ASME B31.1?
o If a pipe was constructed to B31.1, does the boiler code VIII apply in any way to this situation?

• What rules or part of either code relate to the proximity (HAZ – Heat Affected Zones) of the butt welds in question?
o How close by code can two welds be made to each other?

• What relevance does NDT have to assuring that the proximity of the welds is satisfactory?
o Can NDT provide assurance that a design meets code?

 

[metengr]

"The two closely-spaced weldings is NOT a good practice. Generally there should be enough gap (space) so that Heat Treated Regions do not overlap each other."

Absolutely agree with the statement above. There is no reason to cite code or attempt to use the code as rationale for allowing what you have shown in the picture. The bottom line is that overlapping weld heat affected zones is not good engineering practice because of adverse effects on fatigue life and possible toughness concerns. Regarding NDT, if the location of the butt welds interferes with volumetric examination, this is a different matter and would require possible re-work to correct.

Now, can you be forced to remove what is shown in the picture? No, unless the customer specification has a specific provision regarding distance between butt welds for pup pieces at 2xOD, as an example.

 

[Ron]

Agree with metengr...don't overlap HAZ without regard to codes. It's and engineering/metallurgy issue, not necessarily a code issue.

 

[Berndt]

I'm wondering: would it be possible for the OP to determine the amount of strength reduction caused by the overlapped HAZ using something like Holoman-Jaffe? This would then allow design calcs to be redone with lower material strength, possibly proving that the arrangement shown in the pic is acceptable?

 

[metengr]

Kriegfix;
This would take some real effort with conducting metallurgical tests on weld coupons simulating the current weld joint configuration with overlapping HAZ's. By the time you completed this effort, I could have easily welded a new pup piece to correct this situation.

 

[stanweld]

The proximity of the two welds induces high (near yield) tensile stresses across the thin member. It appears that the material is stainless steel and greater sensitization is a certainty. If the pipe spool is in cyclic service or in a corrosive service, reduced life should be expected.

From the picture, it appears that the pipe spool fabricator may have used a repaired flange or fitting, noting the absence of a similar weld on the opposite end connection.
Depending on your contract requirements, you may be able force the fabricator to correct at its expense. Often times contracts provide wording to require prior approval of repairs by welding.

 

[Metalguy]

>"Generally there should be enough gap (space) so that Heat Treated Regions do not overlap each other."<

They appear to be talking about not overlapping weld areas which will be post-weld heat treated, not HAZ's. Their reason is probably related to how extended and repeated PWHT's can reduce strength below acceptable limits.

The subject piece appears to be stainless, so PWHT wouldn't ordinarily be used. What is the process fluid, oper. temps., the type of SS and the C amount?

"You see, wire telegraph is like a very long cat. You pull his tail in New York and his head is meowing in Los Angeles. Do you understand this? Radio operates the same way: You send signals here, they receive them there. The only difference is there is no cat." A. Einstein

 

[moltenmetal]

It looks to me like they hit the dimensional tolerances to make the pipe fit, but missed on the angular tolerance. Or THEY did a fine job, but the flanges that they were mating to were out on angular tolerance. This is common with welded piping and is the reason for allowing field fit welds on piping spools which are shop fabricated (if you actually want the piping to fit!).

If the pipe was shop-fabricated to dimensions given on an isometric drawing, chances are it arrived on site and didn't fit. Rather than sending it back, they repaired it in place.

If welds being spaced too closely was a concern, then the design of this spool was defective- a minimum, adjustable pup piece length between the elbow and flange should have been designed in to facilitate fit-up.

In a situation like this, it's a matter of choice between evils that a pipe fabricator might carry out, often without your knowledge, to make the pipe fit without having to start over.

Would you prefer them to have strained the pipe by the required distance to force the flanges to (sort of) fit?

Or would you have preferred them to heat-strain the pipe to fit (i.e. applying a "diamond heat" to the part with an oxyacetylene torch, often without internal purging while the heat is applied)?

I suppose they could have used long studs and a wafer spacer plus extra gasket to make up the distance, but that would have looked like a screwup.

The option of scrapping a flange, two 90s and a piece of pipe and starting again would probably not have been on offer. No doubt the schedule and budget allowed made sure of that.

Given that the two butt welds appear to be far enough apart to permit radiographs to be interpreted, the owner should count their blessings and move on. That is, UNLESS they stipulated in the bid specs that a particular minimum distance between welds was required, forbidding weld repair etc.

As to the issue of overlapping HAZ: this concern can be valid and if so, must be specified as a concern so that fabricators know to avoid it. Only the designer knows for sure.