Geysering?

Posted by: roveran

Geysering? - 08/23/01 01:08 AM

Nowadays I am concering about geysering.
But I have no idea about geysering.

1) What is geysering?

2) How to calculate in CAESAR II ?

3) What do the informations need for calculation?

Please help me. confused
Posted by: John Breen

Re: Geysering? - 08/23/01 06:32 AM

Hello,

Geysering is defined (sort of) in B31.3, Appendix F, Paragraph F301.5: "geysering: an effect that can occur in piping handling fluids at or near their boiling temperatures under conditions where rapid evolution of vapor within the piping causes rapid expulsion of liquid. In such cases a pressure surge can be generated that may be destructive to the piping. Geysering usually is associated with vertical pipelines but may occur in inclined lines under certain conditions". In paragraph 301.5.1, B31.3 lists geysering as one of the impact loadings that should be considered in design.

We see this phenomenon occur most often in cryogenic piping systems in which product lays stagnant in the line for a while and the temperature get to be near the boiling point. Then when the pump is turned on and the product goes through a valve the boiling occurs and the two-phase flow is pushed into a riser. At that point, the vapor "bubble" expands and an acoustic wave moves up the pipe to impact with the next change in direction (e.g., elbow). The effect is much like a fluid hammer event. I do not know how to quantify the magnitude of the force on the elbow. I suspect that from a C2 standpoint you would model the event somewhat in the same way you would model a fluid hammer event. Rich?

Best regards, John.
Posted by: Richard Ay

Re: Geysering? - 08/23/01 07:02 AM

Yes this could be modeled in CAESAR II in the same manner as fluid hammer. However the unknowns in this problem are very similar to slug flow - you don't know the force magnitude, and you don't really know the timing of the event.

The best thing to do in this situation is model two conditions; a best case and a worst case. One has to be careful not to skew the input to achieve "pleasant" results.