Stainless Steel Threaded Connections in Cryogenic Service

[ToddLoran]

I am looking for feedback/experience using threaded connections on a stainless steel cryogenic system operating at -100° C.
What are the design considerations for this type of connection in this type of service?

 

[EdStainless]

Are these structural fasteners or other?
Will they have to be made (or un-made) while cold?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[GD2]

ToddLoran,
I guess the code for your plant is B31.3. Did you run through Part 4 of the code? Limitations are more on severe cyclic application. It gives some guidance on using threading joints. Usually, you will use threaded fittings in instrument connection lines.

Ganga D. Deka, P. Eng
Canada

 

[moltenmetal]

If it leaks, what happens? Do you lose some inert material, or do people die?

If the latter, you can't use threads.

If the former, it's a matter of finding a thread sealant which will handle the thermal cycling from room temperature to -100 C. You'd need to speak to the sealant manufacturers to be sure. However, if you're going to be relying on just teflon tape, experience shows that teflon tape alone isn't a particularly robust solution for joints subjected to repeated, steep thermal cycles.

Liquid metal (i.e. seal welding) is a very reliable thread sealant...

 

[ToddLoran]

A few additional details:
The design code is B31.3, the connection under consideration will be a 1" diameter thread-o-let welded to a 16" diameter sweet hydrocarbon process line. The TOL is for a pressure sensing line. The temperature cycles will be infrequent, i.e.: plant shutdowns, normal operation will be steady state pressure and temperature.

 

[mk3223]

You need to review the project design guidelines for using threaded connection on the "cryogenic" process if the TOL is allowed for using in the process. Personally, a seal weld or flange connection will be used in this case.

The pressure sensing line should have an isolation valve to separate between process and instrumentation.

 

[Gator]

Another reason is that you might want as little mass as possible for the cooldown bootstrap process. What is the fluid?

 

[moltenmetal]

In non-cryogenic services, the normal practice is to have a root valve welded to the process line any time it's that large, and then to have the instrument attached to that root valve by whatever means are acceptable. A leaking thread is much less of a problem if it's downstream of a root valve which can be closed to allow the threaded joint to be disassembled, sealants re-applied, and then tightened up again.

In a cryo service, the pendant component may a problem from a heat gain perspective, or may freeze up and become blocked etc. We have no idea about the details of your service so we can't advise you further, other than to say that a thread leaking flammables under cryo insulation would likely be a bad thing.