Dear friends,

I am performing stress analysis and pipe supporting for some Flare and steam lines. I have some doubts regarding support and piping arrangements for long vertical lines for instance 7-10meters. Please find the attached file and give your advice in this regard.

In item 1 weight and temperature cause piping deflection in the lower elevation, and this location is a source for the accumulation of condensation. Moreover one or two support me be lift-off due to downward expansion of the line.

In item 2 first support near the vertical run is close to the elbow and vertical expansion is in the upward direction. This causes the lift-off of some supports on the upper level.

Which arrangement is the best practice for flare or steam lines?

Well... I don't think you'll find true consensus here, but I'll offer something.

1. From a generalized standpoint, the best case scenario for stress would be to locate the vertical run to avoid lift-off of supports if possible. That's not to say everyone would appreciate having several risers in the same vicinity with varying distances from the column.

Lift-off can occur due to static effects, dynamic effects, both, or neither.

You might consider whether it's more acceptable to deal with high stresses due to static effects by clamping things down to avoid high stresses due to dynamic effects.

2. Consider structure flexibility. Additional load will be applied to the bottom support from lift-off at the top. Is it naturally flexible enough to settle back down? Can it safely be made flexible enough?

3. Steam: Assuming flow is from low to high, your steam system should have traps to administrate problems away, rather than engineer them away. As such, I'd expect to see the trap on or near that bottom bend.

Assuming they wish you to analyze based on a presumption of slug flow, regardless if there's a steam trap or not, then it doesn't matter which direction flow is in, you'll see a slug load oriented 135 degrees from inlet/outlet of bend at the center of bend. This can exacerbate lift off, or create lift off.

If the natural frequency of the pipe isn't "high," and the slug velocity divided by the vertical run's height is lower than the natural frequency, then the accuracy of a static analysis pales compared to a time history analysis.

4. Relief. The dynamic relief loads act axial to the pipe in the opposite direction of flow as the pressure wave travels down the line. You might consider asking yourself seriously how hot the line should be designed to for dynamic relief loads. The longer a given run is, the higher the dynamic load.

And if it truly is a relief system, and you're looking at this kind of configuration, you should look into the process calculations. The pressure loss in the line is causing the fluid to accelerate, up until the choke point, after which it goes back down again. Therefore, one bend can have a higher load than the next, and it's not guaranteed that the net force is 0.