Allowable Displacement in Vertical Pipe 2

[Bilal7590]

Greetings,

I was carrying out stress analysis of a piping network in case of an earthquake using 'Bentley AutoPIPE' software. The results showed that piping supports were not facing any significant stress, however the vertical section of pipe had a maximum displacement of about 100 mm (4 inches) at a few points.

I would like to know if there is any allowable limit on lateral displacement of a vertical pipe? I have searched for any standard that specifies a limit but i did not find any. Is a horizontal displacement of 100 mm allowable in case of a vertical 6" line carrying a pressure of 25 bar at 100 C temperature?

From whatever I have studied on the internet thus far, my understanding is that displacement is not of concern provided the stresses on supports are not too high. I think that stress on pipe supports is the deciding factor regardless of displacement. Please correct me if I am wrong.

Please note that I am not referring to thermal expansion. I am concerned about lateral displacement due to an earthquake.

Please share your experience.

Regards,
Bilal

 

[pennpiper]

Bilal7590,
Right now I am dreaming up all sorts of configurations, and I am sure none of them reflect what would be your actual problem.
I think you will get better answers and advice if you were to give us an isometric sketch of the actual routing complete with the dimensions of all pipe legs, the location of all supports, guides and restraints along with the location and type of any mass (such as a valve).

 

[BigInch]

For seismic loads, keep the pipe below allowable seismic stress and, as long as the pipe doesn't hit anything else, or slide off the support, seismic displacement shouldn't be much of a concern. Give the resultant support loads to the structural engineers.

If lateral displacement is excessive, you might try adding some guides.

 

[DSB123]

Bilal7590,
Why are you carrying out a pipe stress anaysis if you do not know the limits of displacement? You mention support stresses but what about the pipesystem stress levels - are they acceptable. Are you doing a response spectrum analysis or just applying an acceleration horizontally? How long is the vertical section ?

 

[Bilal7590]

Greetings,
Thanks a lot for trying to help me out.

While conductiong this analysis, I have specified Bentley AutoPIPE software to analyse the system with reference to ASME B31.3 standard. As for allowable stress values, the software then calculates allowable stresses based on the mentioned standard. It then calculates the stress level at each point and displays a ratio of calculated stress to allowable stress. The stress ratio for this piping system came out to be between 0.29 to 0.35. For the vertical section of pipe that I am concerned about, stress ratio was only 0.29.

This vertical section is about 50 meters long and is supported by 2 guide supports and one line stop support.

And I am not conducting a response spectrum analysis. I am applying a horizontal acceleration of 0.35 g's because this is the seismic value for which the plant has been designed.

I have raised this topic because I wanted to know if there is any standard that limits the amount of displacement, because I could not find any.

 

[BigInch]

No. Code limits in that respect are only placed on stress.

 

[Bilal7590]

BigInch,
Thanks a lot for your response.
The reason why I decided to go for this analysis was that this vertical section was reported (by maintenance staff) to be vibrating excessively during an earthquake few months ago. However, obviously it was not possible to quantify the vibration that was detected by the observers.

My initial thinking was that we should be more concerned about stress levels. Hence, I did this analysis just for my personal knowledge and interest; and stress levels came out to be just fine.

Your response has indeed given me more confidence that I was thinking in the right direction.

Thanks a lot once again.

 

[XL83NL]

50 meters long and is supported by 2 guide supports and one line stop support.

That sounds like an awfully poor design for a NPS 6" vertical line. Ever wondered if that causes the horizontal deflection? I wouldnt need AP for that to figure out the problem.

 

[BigInch]

Ya. A couple more guides might be in order, no?

How is it in high winds?

 

[Bilal7590]

Greetings,
Actually, other than these supports, there are 6 platforms for access to different heights of the process tower from which this line is coming out. The pipe goes through the holes made in platforms for this purpose. The diameter of holes is approximately 8 inches; which means that the pipe can move only 1 inch on either side. This kind of acts as a restraint in case of wind as well.
It is worth mentioning here that at the time of earthquake, the observer was at one of the platforms taking samples of process fluid. Otherwise, no one has previously reported high vibration in cases of strong wind or earthquake.
I modeled these holes in platform as guide supports in lateral directions with 1 inch gap on each side.

 

[BigInch]

Using holes in access platforms as "guides" is a bit unorthodox. Not seen that before.

 

[saplanti]

Bilal7950:Yes, you can use the platforms for lateral guiding purposes. However, you need to simulate the stiffness of the platform by modeling at each level. The platforms will move with the tower in both vertical and horizontal directions.

 

[saplanti]

But be careful that the pipe stress software will be locking the gaps (considering zero gap) for the seismic analysis. Therefore you will get zero gap response at each elevation of guides for the seismic load application. This is the way how the available beam element software works. I do not think Autopipe is different.

 

[XL83NL]

@saplanti: doesnt AutoPIPE threat gaps in seismic properly by using it's non-linear analysis, or isnt AP different in this matter compared to CII? Link & link.
@ Bilal: I'd use the 1" gaps on each platform to mount a guide there. As BI said, using the platforms themselves is a bit unorthodox.

 

[saplanti]

I have to make a bit more explanation on the analysis. The previous comment for locking the gap was for any kind of dynamic analysis only. If you choose to use static horizontal loading or factor of g, the gap locking (zero gap will not be available and the lateral forces will fill the gap if they are sufficient. I guess you are using g factor seismic loads laterally, and there will not be a big problem for you to model the guides, and get the responses at each guide level either on the platform structure or on the pipe.

I trust this is sufficient.

 

[saplanti]

XL83NL : I did not see your comment earlier, and you are right for the static analysis. This kind of guides (structural members are attached onto the tower shell)are very common for the tower piping in refineries, and similar type plants. These structural members can be sometimes (or at some elevations) the platform members.

 

[Bilal7590]

Greetings,
In AutoPIPE, one can set the gap for guide supports as explained by XL83NL. And yes it's a good suggestion by XL83NL to add guides in the 1" gaps between platform holes and pipe.

I thank all of you for contributing and helping me out in this scenario. I highly appreciate the kind guidance of you all.

Regard,
Bilal

 

[BigInch]

Be careful using non-designated pipe supports or guides made from holes in grated, or metal checkered plate platform floors. These are often just steel grating or plate with holes, sometimes banded, sometimes not for pipe penetrations. I can assure you that the structural engineers that make that stuff are not aware of what you will be doing, otherwise they would not approve. Often this steel grating is cut completely across through the center of a hole, so as to allow the grating to be pulled when the pipe is contructed. Some grating of this kind is not even tack welded to its frame. Just laid in, for easy removal so you can paint around the pipe, or easily do whatever later. As I said, it is highly unorthodox and such a support likely does not meet piping codes (MSS SP 58). At best, if it functioned as it was supposed to, a good scratch in the coating would certainly be a possibility, if nothing else. There are more appropriate guides that could be used.

 

[saplanti]

Biginch is very true about the load path between pipe and the structural members at support locations attached to the tower . You just cannot leave the pipe without any support with the gap to the penetration. The pipe will move up / down under thermal loads and probably by closing the gap under certain type loading. Therefore you need to introduce adequate support(s) on the pipe and the gap will be between the support and the guiding structural members at each level.

The supports on the pipe can be selected either from a supplier or designed by yourself (to the code) for the purpose. First determine the loads and talk to the support vendors if you need advice on the support types. Please remember that support supplier do not take the secondary stresses into consideration in the selection of supports, they only consider structural calculation to the code. If necessary, working on the secondary stresses is your job.

 

[DSB123]

If you are doing a pseudo-static analysis by applying G values horizontally then the displacements you are getting from the analysis are incorrect. The "constant" horizontal load of 0.35G is not there - in one direction - for a long time. The displacements are dependant upon the natural frequency of the system. By applying the 0.35G load you are checking the stress levels but the displacements are something else. You need to do a response spectrum or a time history analysis.