hello guys,
i am performing the stress analysis of flare header of 92" and two 84" having temp of 270 deg.centigrade.anchor are 168 meter away from each other with one loop of 22 meters.my static seismic factor is 0.34.we are doing static equivalent dyanamic analysis for the whole system.we r getting the huge sesmic load in axial direction in the rage of 721 KN for each pipe.now civil is facing problem to design the members.can any one suggenst how to reduce the loads.i did lots of tries but not come to lesser value.

If you are using static equivalent loads, the only factor relevant to the forces on each anchor is the weight of the piping between that anchor and the anchor on each side. You can analyze until your head explodes and keep coming up with the same number.

The only way to reduce the seismic load will be to reduce the weight between anchors. You will probably want to add intermediate anchors. If this causes your static runs to fail, you will need to use snubbers instead, which can be treated as dynamic restraints but not static ones.

This is not an easy sort of analysis to perform. Keep at it!

.......and it would not be prudent to have the snubbers active for the wind load evaluation (you can't count on them to "lock" on wind loadings). Wind, no snubbers but guide friction is OK. Seismic, snubbers OK but no benefit from friction allowed.

Anil,

The anchor loads could be distributed over other restraints along the 168 meters length.

The first thing to do would add guide restraints along the 168 meters of run, if not already included. The second thing is to add line stops with appropriate size gaps at several of the guide restraints. The expansion loop legs need to be studied for the locations with zero lateral thermal displacement to add guide restaints. If there is not suitable structural steel at those nodes, then gaps will be needed there at the guides. Line stops with gaps could also be added at the loop guide restraints. The gaps would be sized to allow thermal movement to close the gaps resulting in small nominal static restraint loads, and then restraint of the piping for seismic cases.
A more elaborate restraint would be an inclined stop that would be angled for the combined axial and lateral displacements at the support node.

I will have to take another look at Jamnagar site conditions for seismic factors.

The approach I favor is Mr. Yees' in liue of snubbers... gapped guides or stops have an excellent track record as opposed to snubbers...

I solution to these kind of situation I have used is having line stops at suitable locations and just have them gap equal to thermal expansion (Which ever side pipe grows, other side can have 3mm). This way they can hold any excessive displacement due to seismic condition. You will have to twick the gaps in such a way that total load get distributed among line stops and anchors.

Good luck

The load which you have mentioned probabaly includes high thermal force also, if so then line stops are to be loacted at right place where pipe movemnt is very negligible.Seismic loads are normally derived considering operating temperature not design temp.i,e.,W+P+T(ope)+U. Please ensure calculations are done considering operating temperature.Also you may get high loads if the line stop is loacted at the end of pipe where thermal loads are not getting balance by opposite force.

Please send the stress model for my personal interest.

Regards,
Govind Savant
Govind.Savant@akerkvaerner.com

Hello Anil,

Think about to use rollers supports from EU companies like Hidra-Witzenman or Lisega, they have very good performance with a very low cost vs. normal supports.
We used them for a 84" flare line 6 years ago and they are working like a silk. The conditions for our system were, more less, the same that you have but with more seismic factor (0.45). After the installation, the refinery suffered severals seismics events with no consequences (Thanks to God !).

Regards, Roberto

I've been following the development of B31.E (or is it B31.S?) - the B31 recommendations for seismic evaluation of piping systems. This document is quite similar to American Lifelines Alliance document covering the same topic.

They both discuss gaps on guides - you need the restraint for the seismic load but you don't want to harm the expansion flexibility. These docs imply that you can consider the gapped restraint active for static seimic analysis but, if you do, then you should use twice that calculated load (the impact load) in sizing the restraint.

I just looked again at the ALA doc and I do not see a clear statement that, when qualifying the pipe (not the restraint), you can consider a zero gap when running the g load.

It says that as long as the gap is less than 2 inches (on 2 inch & greater pipe), the restraint load can be calculated as if there is no gap.

Anil,

After conferring with structural engineer the seismic load factor of 0.18 to 0.20 was derived using soil class 'C' and response factor R= 0.3 , for Jamnagar site conditions.

Your seismic factor of 0.34 could be high, depending on what soil and response factors used.

Hopefully the expansion loop is located near the center of 168 meters. If the loop is located to one end of the 168 m run, then it could be better to have expansion loops at each end. The center section then could be restrained with a center anchor, plus gapped line stops near each expansion loop.

Has a cut short cold spring been included for reducing the thermal expansion loads?

Anil,
Try modelling the support with structural stiffnesses particularly for guides, line stops. The anchor load will reduce. In such case the ideal way will be composite analysis with structures.

TO Roberto,
'Think about to use rollers supports from EU companies like Hidra-Witzenman or Lisega, they have very good performance with a very low cost vs. normal supports.
We used them for a 84" flare line 6 years ago and they are working like a silk. The conditions for our system were, more less, the same that you have but with more seismic factor (0.45). After the installation, the refinery suffered severals seismics events with no consequences (Thanks to God !).'

Can you explain the reason behind considering roller support for flare piping and whether are you considering double cylinder roller supports as 0.04 friction vertical support only or lateral guide also for all the seismic/wind/thermal analyses ?

regards,

sam

PS: I wish any forum-member who used roller supports in hot piping of similar nature will comment on this topic.

dear sir

iam so have similar problem in analyzing flare line

i use teflon plate (PTFE) for reducing the anchore load in operation and use some snubber for reducing anchor load in seismic


best regards
A.SADEGHABADI

Anil,

Total seismic force in a direction = Total weight times base shear factor in that direction. Approx.load on each restraint= Total seismic force/No.of restraints ( ignoring the effects of Friction).Now you decide what you have to do.Kindly follow Mr.Yee's suggestion on realistic base shear factors.

Understanding mechanics is the first step to solve a problem, which these days ( that is where the old timers have an edge as they did not have computers to do the calculation )most stress engineers lack.


Regards