Dear All,

I am currently working on the analysis of a buried CHW network. The client has requested to qualify the network in accordance to B31.3. [which is very unorthodox for CHW application.]

The network is 750 m (2460 ft). VAL is 60 m.

For B31.3 in the restrained part of the network are failing at all junctions.

If I would analyse this network with B31.1 with axial stress activated the network will pass. To my knowledge, the only difference between B31.1 with axial and B31.3 is the SIF value io ii, but the calculated code stress from B31.3 is 6 times higher than the calculated stress from B31.1 at junctions.

1) can anyone explain to me how is this happening?

I think B31.3 is applying io to axial stress and B31.1 with axial is not. Is that correct?

2) If yes, then is B31.3 overestimating the stress for this application or B31.1 underestimating it?

Thank you.

Amr Elsawy.
Perma Pipe.

Hi there!

Probably, the big differences regarding stresses on intersections are due to the Axial SIFs that are not taken into account by ASME B31.1. You may check if the over-stressed sections belong to header part - if yes, this is the reason...

The immediate solution that comes in my mind is to let the intersections to move - this means to let the pipe un-buried at intersections. Depending on the pipe layout & environment this approach might not always be acceptable.

Another way to act is to review what temperatures you use for analysis - you should use REALISTIC TEMPERATURES, meaning the ACTUAL Maximum & Minimum Temperatures reached by metallic pipe wall. In this regard, see Project Specs, Design Basis and, if nothing is provided there, liaise with Process Discipline - they should provide the temperatures for calculation - might be a tough issue, but this is the situation...

Another direction may be to assess accurate SIFs. Depending on branch/header sizes, the SIFs on the header part may be significantly lower than the default B31.3 SIFs that correspond mainly to the branch part. You may use FEA assessment for this scope - PRG FEA Package r FEA Tools (if you afford this...).

I hope this thoughts may be of some help.

Good luck!

Regards,

Dear Dorin,

Thank you for your reply.

Yes you are correct, that's exactly what's happening.

Unfortunately, I cannot relief the stress by any means other than by using expansion loops. Al the junctions that are failing are located in the fully restrained region on the run. The high intensified axial stress due to lock-down is the main contributor to code stress value.

I think the best way to solve this is by using FEAtools.

Thanks again.

Amr Elsawy,
Perma-Pipe.

Hi All,

I am doing the pipeline stress Analysis as per B31.4 CODE by using Caesar II Basic model in which the stress failure in OPERATING CASE is 114 % but if I do the same system using ALI the stress Ratio is 76 %..... This shows that Caesar II Basic model is conservative. Can you suggest whether my point of view is correct? and which method is good to adopt (ALI OR Caesar 11 Basic model)

Best Regards
K.RAJESH

Amr,

I do not expect an expansion loop in a buried section will give you the flexibility you seek. How can a buried loop expand to limit strain in your tees?

K.Rajesh,

Rather than hijacking an existing thread, it would have been better for you to open a new thread...

In my opinion, the American Lifelines Alliance (I assume that is what you mean by "ALI") would be the preferred reference - it is more recent, more complete and more published. But any great differences should be explainable. I suggest you review the different stiffnesses and different ultimate loads in each.

Dear Dave,

Thank you for the reply.

I'd like to clarify what I meant for you.

I have a long run of pipe 750 m. Virtual Anchor Length is 60 m. Any tee located on this run after 150 m (in fully restrained region) is failing while analyzing with B31.3 because axial stress is huge (compared to bending and torsion) and it's intensified by io.

Same model pass B31.1 with axial stress F/A activated.

Anchoring before or after the tee is meaningless and does not affect the result as displacement is near 0.

Accordingly, if you would model a loop near the junction (less than VAL from the T), axial stress will be relieved because the tee will then fall in the unrestrained region of the pipe.

I made a test model for this. Tee was failing with 140%. When I modeled a nearby loop stresses went down to 70%.

Definitely UG loops are not as effective as AG.

I am not saying that this is a smart solution, or even a solution. But in case FEAtools is not used for SIF I think that this is the only way to relive the stress for this configuration.

Please let me know if you have other ideas, I'd like to hear from you.

On a different note, I'd like to thank you for your work and publications. I have learned a lot from them.

Amr Elsawy.
Piping Design Engineer.
Perma-Pipe.

Maybe if you increase the thickness of the tees.....

Hi Dan,

Yes of course. Increasing Tee thickness relief the stress as well and it is most effective when the cause of failure is axial stress.

Thank you.