UG-36(a)(1) axis ratio > 2x

[chaulklate]

I have a tangential nozzle set high in a horizontal vessel such that the long axis of the opening is greater than 2x the short axis. UG-36(a)(1) indicates when this is the case, reinforcement across the short axis is to be increased as necessary to protect against distortion. Anyone know how I go about confirming this nozzle design? Is there a formula? I can add a repad, or weld build up, or even external stiffeners (not a good idea) for the extra reinforcement, but that doesn't change the axis ratio. And using a heavier schedule won't change anything. So how do I know if this nozzle is adequately designed??

 

[gr2vessels]

This topic around the UG-37 correction factor has been debated several times. Try a word search in the Search Box above for 'hill side nozzles'

 

[chaulklate]

Thanks gr2.

I don't think tweaking the F factor is going to help me here. As it is right now, my area requirement has been exceeded, as well as the nozzle wall thickness. Everything works, except UG-36 says more reinforcement is required...but it doesn't tell me how much reinforcement, or where it should come from. It's quite ambiguous. It says "use more", but doesn't say how much, or provide any guidelines to get the additional reinforcment they want. I'm stumpted.

 

[curtis2004]

chaulklate,
What sizes we are talking about? Can you give more specific?
Thanks,
Curtis

 

[TomBarsh]

chauklate, I suspect the Code committee guys were stumped too when they wrote that provision. They probably knew that something needed to be done, but didn't really know what (in other words, no handy, simple prescriptive rules presented themselves).

The F factor isn't related to this issue. This factor accounts for variation in stress in the shell from circumferential to longitudinal as the plane cut through the nozzle varies its direction from 0° to 90° to the vessel axis. The factor modifies the required area based on that angle, but does nothing to account for the aspect ratio of the nozzle opening.

I was thinking of directing you to the newer nozzle opening analysis methods available in the ASME Code: Appendix 1-10 in Div 1, or Part 4.5 in Div 2. These directly calculate the stress in the shell and are not based on reinforcing aera. But both of these are limited to aspect ratios of 1.5:1! So they are not useful in this situation.

You asked about formulas. I think you must directly calculate the stress in the shell for this condition. There's really no handy closed-form solution for this. I expect that you will have to perform a finite element analysis for this case.

 

[chaulklate]

Curtis - I have a series of 2" 300# Sch160 nozzles in a 42" OD horizontal vessel, offset at 18" both sides of horizontal centreline (3" from top and bottom of vessel). Not that it makes much difference, but the lower nozzle would be slightly more critical due to the static head it is subject too.

Tom - I was thinking FEA all along but hoping to avoid it, thought there might be a work through out there. Shazbat!

 

[curtis2004]

chaulklate,
What design pressure we are talking about?
Why you can't weld your 2" pipe radially, or move it closer to centerline in order to reduce long dimension making it smaller than 2xD?
Regards,
Curtis

 

[TomBarsh]

I am reminded of the time someone had a number of small nozzles located tangential on a large diameter vessel. The openings ended up being like 2" x 20" long slits in the cylinder. Certainly a condition that warranted some extra investigation; I'm sure the basic ASME rules were not intended for such a construction.

 

[chaulklate]

Curtis,
It's only 51 psig design. The nozzles are a series of bridles for level, alarm & shutdown. The locations are critical. We have explored mounting the nozzles radial at top and bottom of the vessel, but we are putting a module into an existing offshore platform and space is beyond tight. Believe me, we have exhausted the easy outs, the nozzles pretty much have to go where they are. We will require client buy-in to deviate otherwise, and I have a higher probability of solving the current global economic crisis before that happens.

 

[jamesl]

We ran into this issue all the time when small instrument nozzles are located close to the bottom of a horizontal vessel. I have never seen FEA or other additional analyses were done for those small nozzles.
UG-36(a)(1) is vague. If you have tangential nozzles of large sizes or with significant piping load, then FEA is justified. I would not worry about those small instrument nozzles.

 

[curtis2004]

chaulklate,
What is material and shell thickness?
curtis

 

[chaulklate]

shell 3/8" SA-516-70N
nozzle 2" Sch160 SA-333-GR6
flange SA-350-LF2 Cl.1

 

[curtis2004]

chaulklate,

I calculated (UG-27) tmin as 0.196" thick wall with 0.55 joint efficiency. If your vessel is reletively short and doesn't experience any other loads and nozzles will hold instruments only, then I would not worry much about 2" pipe tangential nozzle welded to 42" OD shell. It is a small.
What is the medium inside of vessel?

Regards,

Curtis

 

[chaulklate]

The vessel is only 7' long and contains glycol, TEG if I'm recalling correctly.

Thanks, Curtis, I appreciate your research and input. I'm also of the same mindset that I don't need to worry about it relative to the pressure, size, and medium. But, the problem still remains, how do I PROVE that I don't have to worry about it?? How do I convince the end user that they don't have to worry about it. So far, all I have to go on are respected opinions, but I will be hard pressed to put all of your opinions on a piece of paper and have the client accept that. The majority opinion received through here and other sources I've consulted suggest FEA and that is more than likely the route we will go.

Thanks all for contributing.

 

[doct9960]

chaulklate,

To prove mathematically that the 2" intrument nozzles meet UG-36(a)(1), you really need FEA.

If you don't want to do FEA and still want to meet UG-36(a)(1), all you have to do is get rid of the "twisting moment" (or torsion as I interpret the Code). I suggest you support the level bridle from the vessel and get rid of eccentric loads that causes torsion on the nozzles.

What are the external loads on the nozzles and what is the design temperature of the vessel? If thermal expansion is insignificant, external stiffeners or gussets on the nozzles are probably not that bad.

Cheers!

 

[jamesl]

chaulklate,

I am curious why your client would have problem with this issue. Typically, instrument engineers want the bottom level nozzle at 6" above the vessel bottom. The tangential nozzle almost always has long axis over 2X short axis. This happens all the time on horizontal drums. An experinced vessel engineer would just let it go. Your client would be supprised if you present a FEA report on a 2" level nozzle.

 

[curtis2004]

chaulklate,
Even if you order FEA analysis on these two nozzles it doesn't prove with 100% accuracy, because FEA based on assumptions and has limitation. In the end of day your customer might ask to prove your "prove" is 100% accurate.
Situations like this require experience, knowledge, and sound judgement. Sometimes you have to balance between risks and costs, ask youself what if something goes wrong.
You can add 3/8" reinforcement pad, or gussets, or both on location of nozzles if you want to feel more comfortable.
If your customer wants more proof, then you can offer FEA option for more accurate results. But he has to pay for those "more accurate" results.
Regards,

curtis