Flush Patch With Nozzle Repair

[Krausen]

I have a 1-1/2 RFLWN nozzle (304L material) on a tower that began leaking at the nozzle to shell weld about 15 years ago. At that time, they decided to weld circular patch plates around the nozzle after confirming a through-wall crack via NDE. Looking at it from outside, most people thought the nozzle just had a repad, but I confirmed from original design there was no repad provided here & dug into the history to find the patch plate repair notes. This nozzle is also exempt by VIII-1 code from reinforcement as a small opening (shell is 1/4” thk 304L material). Recently the leak began again through the patch plate weep holes. We’ll be shutting down soon to do permanent repair. My question is would you go after just replacing the RFLWN nozzle only to the existing shell or would a small shell flush patch with new RFLWN nozzle be a better/more reliable approach for permanent repair?

 

[pv_mkk]

Looks like, originally, the leakage comes from nozzle-to-shell joint and they go to weep hole.

And if you replace new LWN nozzle same size, can you be able to pressurize the tower to confirm no leakage on that joint?

If yes, then go ahead.

 

[RJW000]

If the leak was never repaired why did it just start coming out of the patch plate weep holes? Was the crack repaired from the inside?

Do you know why it is cracking? SCC? Fatigue?

Anyways, I've had much better success with performing a through-wall nozzle replacement than using a new shell plate with nozzle already welded. You may be able to just excavate the flaw and re-weld the existing nozzle but it depends on the existing damage.

 

[r6155]

Apply air pressure (max 5 psig) from the pad hole and inspect, with a bubble test, from inside the tower (if possible). Then repair the defect and repeat the air test.

Regards

 

[weldstan]

A flush patch shuld have been used or simply repair the nozzle to shell (cracked) weld - not a scab patch with weep holes (reinforcement). Appears to be a thoroughly improper repair.

 

[Krausen]

According to the history notes, the through-wall crack was repaired 15 years ago when initial leak was discovered between the nozzle-to-shell weld. This was done by grinding out the crack first on the ID of tower & root welding with E-308L, followed by successful PT NDE. Then grinding the remainder of crack from the OD down to the root weld & welding out to the OD, followed by successful PT NDE. Circular patch plates were then finally installed around this nozzle on the OD of tower after this weld repair was completed, but it is unclear why this was done in the history notes. After all of this was installed at the time, the tower was then hydrotested to the original tower hydrotest pressure successfully.

For my repair, although hydrotesting would be feasible, it will likely be ruled as not practical, since filling a tall tower with water or distillate & performing a hydrotest will take a considerable amount of time. For this repair, we will likely justify using two forms of NDE in lieu of hydrotesting.

 

[Krausen]

Do you know why it is cracking? SCC? Fatigue?

According to the equipment history notes, the cracking is attributed to Chloride SCC of the shell-to-nozzle weld between the 304L material

 

[Dhurjati Sen]

Do not forget hydrogen bake-out before cutting the nozzle and patch plates.
Carry out a thorough visual examination and thickness survey before carrying out any repairs.
There could be an element of surprise for you.

DHURJATI SEN
Kolkata, India

 

[GD2]

Why don't you directly weld it to the shell with an integrally reinforced nozzle because you got a failure history? It will give you a nice full penetration weld to take more external load, if any. Avoid hydrotest for the headache of the cost involved with it in addition to drying out and the risk of Chloride SCC. Do the NDE as we do typically do for a closure weld.

GDD
Canada

 

[Krausen]

Anyways, I've had much better success with performing a through-wall nozzle replacement than using a new shell plate with nozzle already welded. You may be able to just excavate the flaw and re-weld the existing nozzle but it depends on the existing damage.

Could you elaborate on this? I’m currently going with a new RFLWN nozzle full-pen welded to a 9” dia flush patch that will be welded into a new 9” DIA cut out section of tower shell. My concern with replacing the nozzle only is attempting to weld on the old failed nozzle-to-shell weld that has already been weld repaired once before in its history. Going with a new flush patch will allow new welds area away from the original HAZs & the previously cracked nozzle-to-shell weld.

Tolerance is the virtue of the man without convictions -GK Chesterton

 

[Krausen]

Why don't you directly weld it to the shell with an integrally reinforced nozzle because you got a failure history? It will give you a nice full penetration weld to take more external load, if any.

Regardless of which option is chosen, the new nozzle will be directly welded to the shell with a full-pen weld, along with a 1/4" fillet weld over top. There is no external loading concern on this nozzle in this case, as this is connected to a level gage only.

 

[weldstan]

The original crack was removed and repair welded about 15 yers ago and it was obviously thought that reinforcemnt would provide a longer life, which maybe it did. How long was the vessel in service before the crack was found? Was the cause of cracking ever determined? Assuming the same environment and operational stress levels which caused the original crack may still be present, why would you not expect a second failure to occur over time?

 

[r6155]

@ Krausen
Is this NPS 1-1/2 nozzle the only one with defects?

Regards

 

[Krausen]

@ Krausen
Is this NPS 1-1/2 nozzle the only one with defects?

This nozzle is the only leaker historically & only nozzle with known weld defects at this point. The tower has been routinely API510-inspected internally since it was constructed in 1985.

This is a level gage nozzle at the very bottom of a 20’-tall tower in diesel service (predominantly gas phase, but in liquid phase at this elevation of tower). There are no other nozzles located lower in elevation on the tower shell. If Chloride SCC really is the issue, I’m wondering if this tower isn’t seeing some condensed water vapor at this specific location at times or regularly. But I’ll need to check with the Process folks on that.

 

[Krausen]

The original crack was removed and repair welded about 15 yers ago and it was obviously thought that reinforcemnt would provide a longer life, which maybe it did. How long was the vessel in service before the crack was found? Was the cause of cracking ever determined? Assuming the same environment and operational stress levels which caused the original crack may still be present, why would you not expect a second failure to occur over time?

Valid questions, but the tower was in service for 18 years before this nozzle leaked at weld. I don’t think we know enough to justify a material upgrade of entire tower from what the current failure cause theory is at this one specific nozzle (Chloride SCC). The reason for installing circular patch plates around this nozzle after weld repairing it initially was not explained, but I’m glad they did, as the subsequent leak in 2020 was found through the tapped 1/4” DIA weep holes on these patch plates, which we used for installing threaded 1/4” valves to in order to stop the leak.

 

[r6155]

@ Krausen
I think the problem is the welder or the welding procedure.
I suggest:
1) Make a mock-up with the same NPS 1-1/2 and plate 1/4" in the same position as on the tower. No reinforcing pad.
2) Then NDE.
3) Try to make a defect like as real one on the tower, then NDE
4) Repair the mock-up in the same position and same welder. NDE.
5) Repair the tower keeping all conditions, NDE, air test with the pad and bubble test from inside of the tower.

Try to witness and take a video

Regards