High Temperature PSA

[FDS2008]

I am working on a project that involves a short section of piping that is at 1,800°F and 5 psig design conditions. I have been requested to provide the stress analysis. They want the design to B31.3, even though it is not considered pressure pipe. The conundrum is what allowable stresses to use as B31.3 only goes to 1,500°F and the stress is only 1,100 psi. The material they plan to use is A316-Type 321H. I have a similar question for the flanges (B16.5 allows 1,500°F @ 40 psi).

I know that B31.3 considers high pressure design but what is your experience with high temperature piping? Any suggestions as to how to approach this from a pipe stresser point of view?

Thank you in advance!

 

[TGS4]

You are getting well into the creep regime with those temperatures - even the self-weight will have a significant impact.

I'm not sure that such a job could be performed using only typical pipe stress analysis tools.

What is desired life? What are the operating conditions? The operating conditions are more important in this scenario than the design conditions.

 

[metengr]

If you intend to consider this a B31.3 design, you will need to look at alternative materials to ensure selection of an allowable stress value. Maximum permitted temperatures are stated to ensure reasonably safe conditions. If you intend to use B31.3 as a guide, this is a different manner and now you have some flexibility to look at more detailed analysis, should you use the stated material.

 

[XL83NL]

I was thinking about API 530, since its also for the high temp stuff, but the scope and application of that standard is different from yours.
You can have a look at it, maybe there are some general guidelines in there, but I dont think it's design code for your application.
I guess youll need due consideration of basic mechanics, added w/ creep, and sound engineering practice on this one.

 

[weldstan]

Assuming air exposure, you will need more chromium to resist oxidation than you have with 321H; however, the service life design may be short enough to use.

 

[FDS2008]

Thanks all for you input!

I have just been informed that the max pressure (rupture disk burst pressure) is 30" w.c., so pressure will not be an issue, which works in our favor.

The operating conditions are close to the design: Max 1700°F and normal 1600°F. I can use B31.3 up to 1500 °F but the material needs to be heat treated to a min. of 2000°F as indicated in the note in Appendix A to use these allowables. Their spec calls for heat treatment to 1625°F, once again against us. The normal ope temp is also 100°F above the max B31.3 limit.

I understand fully that this will not be designed to B31.3, but I am using it as a safety guide to have some peace of mind that this will be a safe system. If I don't get a good handle on this, they will need to retain an engineer that does have experience with this.

If the design was requested per B31.3 we are looking at a 20 year life span, if I am correct.

Are there any calculations or standards that account for creep, to de-rate allowable stresses and what to do about the extra 100°F (fm 1500 to 1600) of heat w.r.t. selecting allowable stresses.

 

[weldstan]

20 year life for 321H at 1600F to 1700F? One can dream.

 

[TGS4]

At those temperatures, you will need to address more than what is covered by a typical B31.3 approach. Plus, you called this a PSA (pressure swing adsorption), which to me also means cyclic service. So, combining cyclic service (fatigue) with those high temperatures, I think that you are going to need to call in some experts.

I have been involved in extremely high-temperature designs (2000°F +) with exotic materials, but even then we were using a design-by-analysis approach with self-derived allowable stresses, which limited to life to much much less than 100,000hrs. Based on what you have written, I would tend to think that the piping that you describe will need a similar approach.

Or, you can line the pipe with refractory.

 

[XL83NL]

I thought of PSA meaning pipe stress analysis

 

[FDS2008]

XL83NL is correct, I was referring to pipe stress analysis.

How would the refractory improve the situation TGS4?

The piping is NPS-4 Sch 40 by the way and is actually exhaust piping from an activator through a HEPA filter and to atmosphere.

If it is a time relates issue, then perhaps they have to replace the piping every 5 years. Not the greatest design but it is their design.

 

[metengr]

There is so much more information required as to the contents of the gas stream besides operating temperature. Again, the 321H material is limited by several factors, mostly time dependent damage in the form of creep deformation and oxidation from corrosion. Selection of refractory may be an option to reduce the mean metal temperature of the pipe wall and reduce exposure to oxidation. However, lining of a 4" pipe with refractory will reduce the ID to a point where it may not be feasible for operation.

 

[XL83NL]

Would it be an option to use a higher pipe size for refractory inning, such to reach an ID approx equal to 4"/40S after lining?
And would 310 be a better option against oxidation?

 

[metengr]

The same upper maximum temperature limit exists at 1500 deg F. For a maximum of 1800 deg F you are in Ni-base, SB-622, or Haynes Alloy 230,

http://www.specialmetals.com/documents/Inconel alloy N06230.pdf

 

[TGS4]

Ah - that clarifies things a bit more.

Considering that you are only discussing exhaust piping, then I would definitely recommend refractory lining of the pipe - just as XL83NL suggests. That would greatly reduce the design metal temperature back into the B31.3 temperature range.

 

[FDS2008]

Thanks all for your great replies. My client just communicated that they have carried out this design before and although they prefer using Incoloy 800 HT, they have designed plants using A312-321H that are still standing after 30 years. Perhaps the actual operating temperature is way lower than what is printed on paper.

The new scope for them is their client requesting a PSA report which is how I got involved.

I recommended to model the piping to the highest temperature below which creep will not be an issue. I understand that this is approx. half the melting point (to be confirmed), although I read another report that indicates that this is approx. 1050° for SS. This is as far as I am prepared to stick my neck out and if it is contended with their client, we can recommend to change materials to Incoloy.

Any other comments please let me know and thanks again for the knowledge.

 

[XL83NL]

Even 800HT (let alone inconel X or 718) wont keep it at those design conditions. The operating conditions had to have been way lower in order to keep it out for 30 years, I guess.

The (approximate) onset of creep can be found in ASME II-D. From the top of my head, itll be around 600 C for 321H.
1050 deg C is where SS is typically annealed.

Anwar, as TGS4 posted stick to linde pipe. B31.3 has a special section for this, so the code 'will suit you well'. For the shell, you can then switch to e.g. standard A106B CS Pipe, which is way chapter than 800H/HT.
Good luck.

 

[FDS2008]

I meant "stress it to 1050°F", which is half the melting point (according to this article -

http://www.nationalboard.org/Index.aspx?pageID=181)

of A-312-321H, at which I do not need to be concerned about creep.