Swivel Elbow Expansion Loop

[Ltorrado]

Hi,

I am currently designing a "special" expansion loop for the well kill line of an offshore platform. Well kill line comes off wellhead and is about 30' away from platform. According to a wind load analysis made by our civil department, we can expect wellhead displacements of up to 2.5" and 1" (laterally, X and Y axis). We can also expect 9" growth(vertically, +Z axis). The vertical growth is due to high temepratures coming off the well (400ºF at the wellhead). I am operating in extremely high pressures (30,000 psi).

I've been asked to design an expansion loop using a combination of swiveling elbows to allow for the vertical growth AND the lateral movements. I came up with a "scissor style" arrangement of elbows. Now, here are my questions:

1) Material to be used is AISI 4340 (120 ksi yield) which is a forged steel rod that is drilled through. Material is not listed in Caesar's database, that's all the information I have. But I looked for several equivalent ASTM numbers and I came up with A322, A331, A752. The problem is that none of these are listed in B31.3. So can I use a material that is not listed by B31.3 and still use the B31.3 calculation method?

2) Has anyone ever dealt with these swiveling elbows? I suggested using an expansion loop but manager wants to stick to these. Should I be worried about leaks developing as time goes by? Thermal growth will put most stresses, but we only expect it during start-ups.

I would appreciate any help.

 

[KernOily]

Well Caesar itself only cares about nu, k, E, and Sc. As I see it, the Sc is going to be the problem since this is not a 31.3 listed material. There is probably a Code case someplace that addresses this very issue but I don't know for sure. How does one research the Code cases?

I have experience with swivel joints in steam service (1100 psig and 550 F). I can tell you they always leak eventually and they have to be maintained. So you will need to plan for the joints to be 'inspectable' and accessible/maintanble by inspection/maintenance staff. For those reasons we try to avoid them. I would try to sell your manager on those attributes. The Ops guys are going to curse you all for years to come every time they have to maintain those dang swivel joints... ;-)

 

[BigInch]

Just what we need, swivels at 400ºF & 30 KSI. Offshore yet.

"I am sure it can be done. I've seen it on the internet." BigInch's favorite client.

"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.live.com

 

[Ltorrado]

Thanks KernOily,

You make a good point and I already had these things in mind. Maintenance is definitely an issue. Accesibility is a problem because this line will be below the bridge connecting the wellhead to the platform. But like I mentioned, we expect the thermal growth only during start-up (once in a blue moon) and the lateral displacements are for hurricane force winds.

I am trying to keep the bending stresses on the elbows to a minimum to avoid development of leaks over time. Will this be enough or, typically, do these valves leak no matter what design?


Regarding the rest... according to B31.3 par. 302.3.2(d) calls out allowable stresses for "other materials". I will be using these to input properties into Caesar. I am working on getting the Sy at temperature, but it's proven to be a difficult task. I am talking to the vendor tommorrow to see if he can give me some information.

BigInch,

I appreciate your reply, but instead of making a smart ass comment, you could share some experience you've had with this type of situation. That would help me a lot more.

Thanks.

 

[KernOily]

"I am trying to keep the bending stresses on the elbows to a minimum to avoid development of leaks over time. Will this be enough or, typically, do these valves leak no matter what design?"

There is a minimum moment required to break the joint loose, i.e. to make it swivel; a 'break-away' torque. That value comes from the manufacturer. So you will take the moments calculated by C2 and check that against the required minimum to see if the joint wil rotate. There is a modeling technique in C2 to properly model a ball joint so make sure you look that up. I think it's in the 'Technical Reference' section. There will be some residual moment present on the joint after it swivels because the swiveling relieves some of the moment but not all due to the friction present in the packing. Again, I would do this design work in careful conjunction with the joint manufacturer. Chik-San and Barco are the two I am most familiar with.

As for your Sy, why can't you use the nominal Sy (specified minimum yield strength) from the steel manufacturer, or else the AISI 'book value' for Sy?

Again, in my experience, yes, these things always leak. In steam service the packing is maintained by using a grease gun. You inject high-temperature grease into the joint to maintain the seal. Even then, they still start leaking after a couple years. I would guess your service is no better in that regard. Again, I would make sure you are doing this in close conjunction with the manufacturer. Also I would post this question over in the offshore forum and the petroleum production forum to see if anyone there has any experience.



Hope this helps...

 

[Ltorrado]

Yes, you bring up another good point. I am already familiar with the COADE ball-joint article talking about static friction torques. I am trying to obtain the break away torque from the manufacturer. I already modeled them as rigid pieces with cnodes; I will have to put stiffnesses in later.

I have the Sy already, but not at 400ºF. I believe the yield point will decrease linearly as temperature goes up until a certain limit (somewhere above 400).

 

[KernOily]

Well 4340 is carbon steel, so a reasonable first appproximation would be to curve-fit the yields from something like A106B, or some other steel that is metallurgically similar to 4340, for which there is a Sy vs. T curve (table of values actually) in Appendix A of 31.3, and then apply that same curve shape to 4340. That way you can get defensible answers to keep your work moving forward til you get the real Sy's you need. Man I am just a helpful dude today ;-)

 

[Ltorrado]

Yes you are. Thanks man!

 

[KernOily]

Right on. Please do keep us updated on this, as I want to hear what you find out on the Sy issue. Thanks!

 

[BigInch]

I did.

"I am sure it can be done. I've seen it on the internet." BigInch's favorite client.

"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.live.com