Pipe thickness requirements for high external pressure

[mechengr12]

I have a nozzle inside of a vessel that is no longer used and needs to be removed or isolated to remove the pocket it creates that can fill with hydrogen. The setup is a 8" Class 1500 nozzle on the outside of the vessel with a Sch. 80 8" nozzle on the inside of the vessel approximately 5" long with a class 150 flange. I would like to cut out the internal flange and weld a cap in its place. The issue I have is that the pipe/nozzle internally needs to be rated for an external pressure of 1749psig at 650F. The material is C.276 SB575. Sa is 24.6 ksi at elevated temperature. How thick would the pick need to be for the external pressure differential being 1749psig? I have stepped through UG-28 calculation and come up around 1". Hoping someone might be able to verify. OD of the pipe is 8.625. The inside of the pipe will be open to atmosphere.

 

[StoneCold]

mechengr12
I am doubtful that it matters if an unused nozzle fills with hydrogen gas inside of a reactor from a process or safety standpoint. However if it is bothering you why don't you leave the nozzle there and fill the space inside the nozzle with a short pipe stub or machined solid material. Then the volume of the dead leg will be drastically reduced without modifying the vessel, just the blind flange on the unused nozzle. I am curios as to why you want to do this???

Regards
StoneCold

 

[mechengr12]

The pocket of hydrogen gas poses a huge safety issue. Years earlier a pocket of hydrogen remained in a nozzle on the top head even after being flushed with water. The issue was this hydrogen was not vented from this nozzle located on the top side of the head. Hydrogen takes very very little to ignite. While unbolting the flange, the energy from unbolting was enough to ignite the hydrogen and blow the blind off of the vessel. The incident caused extreme facial damage to an employee. There is a chance in the future someone would wish to remove the pressure blind from the outside without taking consideration of a potential for hydrogen buildup behind the blind.

 

[SnTMan]

mechengr12, I do not find your material. SB-575 appears to be plate/shee/strip w/UNS numbers Nxxxxx not Cxxxxx

Can you confirm? Also confirm 5" length, 8 5/8" OD.

Regards,

Mike

 

[mechengr12]

SnTMan thanks for your reply. MAterial is UNS N10276. OD is 8.625". Length from vessel wall to flange is approximately 5". I have attached an image. If you have any good alternatives that wouldn't result in great cost and vessel modifications it would be greatly appreciated. The vessel is CS with a 304SS liner.

 

[marty007]

I just ran a quick calc and came up with a similar thickness (~1" pipe)...

Given that you have a sch 80 pipe (1/2" thk), you would also have to remove the internal projection. In this case you might as well just weld a S.E. head directly on the inside surface of the shell, doing away with the pipe completely.

What about going in the other direction? As I understand, the safety concern revolves around someone unbolting the blind. What if you were to go the other way and put a pipe cap on the external pipe? The hydrogen would still be there, but you are removing the ignition source...

 

[SnTMan]

mechengr12 getting 0.818" for the nozzle, 1.045" for a head, both from plate, although not likely different if pipe. Excludes any CA.

CS w/ 304 liner and Ni-Mo-Cr plate nozzle?

Not sure what to suggest, other than flush patch.

Regards,

Mike

 

[wailunwu]

Can you cut out the nozzle at the shell and replace with insert plate (same thickness and material as shell, full penetration butt weld, full NDE)?

 

[marty007]

At that pressure, and from the picture attached, that must be a very thick vessel (4+ inches). A flush patch would probably be very difficult at that scale. You're probably still best to do something with the existing opening (either externally or internally).