change of flange

[picasa]

On one of the nozzles on a horizontal drum fabricator put a 150# flange instead of 300# (By mistake).
The vessel is already stress relieved by not hydroed yet.

The nozzle has to have 300# flange. What should be done??

Torch the flange off, put the new 300# flange on and then local post weld heat treat ? How does that affect quality of vessel ? Is this allowed by code? ( ASME SEction VIII Div 1)

 

[tickle]

If I was the purchaser of the vessel I would want the flange replaced and the whole vessel re-stress releived.

Put it down to experience.

 

[SnTMan]

In my experience:

If vessel is carbon steel, P1, you should be able to local.

Any NDE required by Code or customer spec would have to be performed.

If you are not the owner, you should get owner's approval before proceeding, as well as your AI.

If vessel is another ferritic material, i.e. P4, the above applies as well, but simulated PWHT must be performed on vessel materials on the total number of cycles of PWHT the vessel will see, often including extras for field repair (usually required by customer spec). Standard parts such as ASME flgs can be exempted, but pipe, plate, custom forgings cannot. See UCS-85. An exception exists for annealed matls.

From the above, ASME apparently believes that several PWHT cycles has no effect on CS materials, but does on lo-cr materials.

Hope you got carbon.

 

[jte]

picasa-

Cut the flange off, put the right one on... and then decide whether or not you need to PWHT.

You didn't state why the vessel needed PWHT: Is it due to service or due to material thickness? If PWHT is req'd due to service then perform a local PWHT and move on. Re-cooking the entire vessel would provide no benefit and would be strictly punitive. From UW-40(f): The term nominal thickness... is the thickness of the welded joint as defined below. For pressure vessels or parts of pressure vessels being PWHT in a furnace charge, it is the greatest weld thickness in any vessel or vessel part which has not previously been PWHT. If you had to PWHT the entire vessel every time you put a new weld on then you would never do a local PWHT for a repair.

If the vessel was PWHT for material thickness, then chances are that the nozzle to flange weld is less than that which would require a PWHT. Take a look at Table UCS-56. If you have a P1 material such as SA106 then as long as it is less than 1.5" thick you have some options. Take a close look at Note 2 for that table to see where you stand. Remember, it is the weld thickness, not shell thickness which governs. If you have a weld neck flange, then the nozzle neck thickness is what you'd be looking at [ref: UW(40)(f)(1) OR (f)(5)(e)]. If it is a slip on then chances are that your throat (not leg) dimension will easily qualify for the exemption [ref: UW(40)(f)(3)].

Bottom line... This should not be a big deal, and at most require a local PWHT.

jt

 

[picasa]

Actually the nozzle is RFLWN and not a weld neck. Do anyone have any additional comments?

 

[SnTMan]

Well, that may complicate things some. Is it set on the vessel wall or inserted?

I guess you could cut the neck and buttweld it, but that's kind of nasty. Definitely would get owner approval before doing that.

 

[jte]

picasa-

You're just teasing us with the details! Let's have some more like: Why is PWHT necessary for this vessel to begin with? What nominal size LWN do you have? How thick is the shell? Or is it in some kind of head? What materials (might be inconsequential, but it doesn't hurt to let us know)?

As SnTMan pointed out, having a LWN may complicate things. The set-in vs set-on may be critical for determining weld thickness: Set-in with a 3" thick shell may be entirely different from a 2" LWN set on with a roughly 3/4" weld.

But... the ultimate answer still lies with determining the weld thickness per the sections of the Code to which I referred you previously. Unless the vessel is PWHT for process reasons, not Code. If you do wind up requiring PWHT for whatever reason, a local PWHT will be fine.

From my perspective as an owner's engineer, I don't see that I would have any direct control over this situation. Would I like a courtesy notification? Sure. But I issued a PO for a Code stamped vessel which meets our standards. And fixing fabrication errors before the U stamp goes on is entirely code legal and not covered in any of my current standards nor in any of the many client standards I worked with in my past life. Yes, there are some exceptions, but those are usually found for specialty materials (alloy, clad...). As long as your contract (including their standards/specs) with the client does not specify that they shall be notified of any fabrication errors which have been fixed, I do not see any requirement for picasa to notify the client.

jt

 

[picasa]

gentlemen all your input is very much appreciated. I apologize for not providing enough info earlier.

The vessel has HIC plate. The service has some H2S and hence PWHT is reqd for process reasons. The nozzle is 4" LWN. It welds on a 1/2" thk plate.

 

[SnTMan]

For 1/2" CS, I'd cut the whole thing out, replace per original design, whether set on or inset, local PWHT and be done.

jte, most of my work is done for E&C companies, usually very detailed specs, outside inspection, and so forth, no contact with the owners at all. I agree, notification may not be required, I prefer to do so anyway. All you need is for the outside inspector to notice something unplanned going on and all of a sudden there are all kinds of conference calls and stuff, the inspection can get very picky. It is possible to wind up with a perfectly servicable vessel you can't ship.

It's not worth the risk to me. Trust is too hard to build and too easy to lose.

 

[weldtek]

I agree with SnTMan as far as keeping customers in the loop as to repairs or changes. Most ' good ' shops have an internal process to handle non conformances, like generating an NCR (non conformance report) on which they detail the problem, and a proposed corrective action. Depending on the nature and severity of the problem, in addition to internal signoffs, this may be signed off by the customer and in certain cases by the A.I.
The last thing a customer wants to feel is that the fabricator is hiding or covering up something. Anyway that's a call each shop must make.
As far as cutting off just the RFLWN flange and replacing it, we have had to do so on occasion, without problems. A portaband saw will minimize cleanup of the pipe neck, and it may simplify the PWHT.

 

[jte]

Folks-

I have no problem keeping the customer in the loop! After all, I am one (but with previous experience in the E&C world). All I was pointing out is that it is usually a courtesy notification, not written into the contract and certainly not a Code requirement.

picasa-

Since you are PWHT for process reasons, there is no question that you need to PWHT this repair. A local PWHT will do. As I see it, you have three basic options: First, as suggested by the previous posts, cut 'er out and replace with the CL-300 LWN. This would be the cleanest option. Second, consider that the wall thickness is clearly not an issue with this nozzle (1/2" shell vs 3/4" thick CL-150 LWN and 7/8" thick CL-300 LWN). If this would be a difficult configuration to local PWHT (interferences etc) consider discussing a minor design change with the client: hack off the nozzle neck a few inches from the shell and add the CL-300 LWN to the stub. There will only be a 1/8" change to the radius. Use a proper thickness transition etc. This is a bit messy, but if you have practical problems with a direct replacement, this approach could work. You may have to revisit nozzle reinforcement. If that's the case, let me know. I have a few tricks up my sleeve on that issue also.

weldtek-

Welcome to the forum! Sounds like you'll be providing a valuable perspective. I'd suggest you take some time to fill out the "personal profile" box (link on the top left of the screen). If you click on the "tickle", "jte", or "SnTman" at the top of our posts you can see ours. (You too, picasa!)

jt

 

[SnTMan]

It's so cool to have a brief discussion in a forum like this, and reach some consensus about solving a problem. How come it doesn't work that way at a "real" company?

picasa, let us know how it all turns out.