2 replies to this topic

[rajbhar_s]

Hi all,

 

I am rating a rupture disk for a flashed two-phase flow with 0.1% vapor manly consisting of methane and some CO₂. Per API 520-1 Annex C, if I have 'some' non-condensable gas, I can use Direct Integration Method or Omega Method for calculation of two-phase relief area.

 

I  used both Liquid-only flow (API 520-1 eqn 33) method and two-phase flow method (API 520 Part 1 eqn C.10), and as per the results, the calculated liquid-only area is 10% higher (i.e. more conservative) than two-phase calculated area. 

 

My question is,

1. Is it Direct Integration Method or Omega Method is a correct choice for this application (i.e. 99.9% liquid or 0.1% vapor)?
2. What does API mean by 'some noncondensable gas, is it 0.1%, 1%, 5% or 10%? Do they have any guideline?

Regards,

Sanjay

 

Attached Files

•  image_2023-11-28_185358382.png   232.58KB   0 downloads

[latexman]

Is that 0.1% by weight or volume?

 

1. Either is applicable.
2. From a practical standpoint, when a component, or a group of similar components, are in the 1-10% by weight concentration range, I start to wonder if they will truly affect my estimates, or not.  Sometimes they do, and sometimes they don't.  Until you have the experience, you have to calculate.  At 0.1% concentration, I wonder if that is so low it can just be ignored?  Your identification that the area of the liquid case > 2-phase case seems to prove this out.

My $0.02.

[PaoloPemi]

the main difference (Omega vs. Direct Integration) is that Omega adopts a simplified model (see Leung ""A Generalized Correlation for One-component Homogeneous Equilibrium...") , as general rule the Direct Integration procedure should be more accurate / reliable  ...
I would verify your result  "the calculated liquid-only area is 10% higher (i.e. more conservative) than two-phase calculated area"... normally the opposite is true (at least  for common compositions /  operating conditions)
there are many  models available for flashing chocked flow (to mention a few : homogeneous equilibrium, homogeneous non equilibrium, models including corrections as boiling delay etc. ) and, as you may imagine, each model has some specific range of application,  
while I utilize preferably HEM (the version included in Prode Properties) which gives normally conservative results, it is certainly useful to compare different models when the value of some parameter / condition is outside the normal range... 

as example see the different results obtained with a small amount of vapor at PSV / RD inlet...