After a pumping is completed through this 3inch diameter line, which has uni-directional segments as long as 300 ft, then 40 to 70 psig nitrogen is turned on through a 3/4 inch ball valve. A slug flow situation follows that causes the line to rock & jump.

The line has one expansion loop & is anchored to a few points that tee off of the main line.

The line is on hangers and shoes to accomodate for thermal growth. Temperature can vary 150F on this 316 stainless line.

Looking for methods to safely dampen the pulses.

The line is filled with product that must be blown out to the destination tank.

Thanks[*]

[ October 10, 2001: Message edited by: tremont ]

Hello,

Whenever I think of "slug flow" I immediately think "drain it". If I understand the description of the problem, the pipe is partially filled with product when the nitrogen is introduced. The nitrogen flows through the pipe at a fairly good velocity and it picks up a "puddle" of product (as a slug)and carries it in the direction of nitrogen flow. When the slug encounters any (or every) change of direction (the inside face of the pipe wall at a change in direction) it impacts the pipe wall transferring the flow energy to the pipe - BANG!!! I think you have a choice of thoroughly draining the pipe so that a "slug" cannot be formed or significantly slowing down the incoming flow of the nitrogen. The slug is formed when the "puddle" of residual product accumulates until the surface area reaches a "critical" size. The flow of gas over the surface of the "puddle" causes friction and when the "puddle" becomes large enough that area is great enough the friction allows the flowing gas to "scoop-up" the "puddle" and create a slug. Of course, if the entire pipe is partially (and contiguously) filled with product, slugs can form several places at once.

So, I think the cure is not in damping the energy (a typical approach to true "fluid hammer"), but rather in stopping it before it can begin.

Any other thoughts from my esteemed colleagues?

Best regards, John.

Hello Phil,

As long as there are liquid slugs being carried along the piping by the flow of nitrogen, you can expect the piping to take a beating. "Plan B" would involve the installation of "knock out vessels" ("drums" or "pots") which would remove the liquid from the stream by impingement or centrifugal separation. The knock-out drums would have to be solidly anchored (which implies that more expansion loops would be included). A variation of this scheme would involve in-line baffle chambers, each made from perhaps, NPS 8 pipe, with a labyrinth of baffles (somewhat like an automobile muffler) and with a drain pipe for the separated liquid. The chambers would be about a foot or so long and would include B16.9 reducers at each interface with the NPS 3 piping. You would have to solidly restrain (anchor) the piping at these chambers to take the impingement forces to the ground (again, more expansion loops). Of course, the knock-out drums or chambers would remove the liquid but it would have to be collected and pumped to the terminal point (tank). I recognize that this scheme might add an unacceptable degree of complication to the system. Maybe it would be possible to slope ("pitch") the NPS 3 line back to a collection vessel (continuous nitrogen cover) and pump the liquid to the terminal point ?

In piping where OCCASIONAL steam hammer events (e.g., SRV blow-down events) are expected, mechanical snubbers (the hydraulic type are maintenance intensive) are included in the design. I personally do not like to employ this method because designing proper pipe attachments is a concern (I have seen pipe form a plastic hinge at the attachment clamp in a fossil fueled power plant due to the violence of the impact event - total load rejection turbine trip). We sometimes design "doubler plates" to be attached to the pipe locally to spread the load over a greater area and effectively increase the pipe thickness at the point of snubber attachment.

You can see that the "stop it before it begins" approach is the best (most simple), but if it is not possible some rather creative (and perhaps expensive) alternatives may be necessary.

Still looking for additional input from our esteemed colleagues!

Best regards, John.

Hello,

Does appear this phenomenon at pump start-up? It should be, because the fluid speed is greater in that case then at pipe blowing with nitrogen. If this is real, you have to eliminate the hanger supports and replace them with shoe with displacement limits in the direction of pipe jumping. In any case it could be useful a piping system analyze in dynamic case to see the optimal points of piping limits stop. I should like to help you if you want that.

Best regards,
Dragos

I am still working to solve this problem. You are certainly correct
the problem does begin when the pump starts & it is severe. I am looking at the problem using Caesar & also Nayyar's piping handbook. Once I get improve my knowledge base I will be more succesful at communicating.