10 replies to this topic

[tdeb]

Hi all,

 

I would like to know if anyone has ever been able to analize the reliability of Hysys' calculation of pressure drop for two phase vertical lines. In the case I'm working, I have a line that goes up 40 meters, and in the top part of the line, the mass vapour fraction is as low as 5% (propane in liquid-vapour equilibrium). My first guess was that given the low mass fraction, the pressure drop given by height change would be similar to the one that a line full of liquid would have, which was about 2 kg/cm2, so I was expecting a pressure drop of at least that amount (without even considering friction pressure drop that would add even more to it). To my bewilderment, Hysys calculated a total pressure drop of about 0,4 kg/cm2g. Trying to understand what was happening, I was surprised to see that in spite of the low mass fraction, the actual volumetric fraction of the product at the top of the line was about 90%. In hindsight, I guess that is expectable given the difference between both phases' density (570 kg/m3 against about 3 kg/m3). The problem is, it doesn't seem right to me for the potential energy changes to drop to almost nothing with such a little mass of propane evaporated. Even if the volumetric fraction of the vapour phase is so high, the amount of area occupied by it would be dependent on the velocity of the phase. Furthermore, the flow pattern Hysys estimates is dispersed bubble which, I gather, means that most of the cross-sectional flow would be occupied by the liquid.

 

Looking into the pipe segment calculations, velocity phases calculations don't seem to account for the actual fraction of the cross-sectional area occupied by the phase, rather, it calculates the superficial velocity (which would be volumetric flow / total cross-sectional area).

 

To give further information, the model I'm using is the Tulsa 2-phase. I tried other models though, and all of them calculate almost the same pressure drop in the vertical pipe.

 

If anyone has any information that could help me understand this better or convince me that Hysys calculation is correct, I'd be really grateful.

[Bobby Strain]

I have always used Duckler constant-slip model for both horizontal and vertical flow. With Humark holdup. My website contains the calculation software. Flow regimes for Baker and Aziz are shown, too. I have never been convinced that static pressure loss is additive to the dynamic loss. If you wish to be conservative, add the static head calculated with the in-situ density.

 

Bobby

Edited by Bobby Strain, 19 June 2023 - 06:24 PM.

[latexman]

The potential energy due to elevation is usually conserved; not lost like friction. Refresh your recollection of this basic concept with Bernoulli’s equation.

[tdeb]

Dear latexman, 

 

Thanks for your answer, however, even if what you're stating is true, it would apply if the 40m I'm saying the fluid in my system goes up, would then go down again, but that is not the case. I'm stating a system with an initial height of 0m and a final height of 40m, so there should exist a difference in pressure due to static pressure loss.

[tdeb]

Dear Bobby, thanks for your insight. I don't get, however, what you mean when you state "I have never been convinced that static pressure loss is additive". If you could explain further, I'd be grateful.

[Bobby Strain]

To clarify, once you have determined the dynamic pressure loss that should be the total. Static head is not additive. You should do a bit of research and report back.

 

Bobby

[latexman]

Static pressure loss is not strictly affected by flow rate. Static pressure loss is the same at zero/low flow as at high flow, so it will subtract out to no net affect on frictional pressure loss.

 

I still sense some confusion over pressure drop calculation and mechanical energy balance.  What is the purpose of  the pressure drop calculation result?

[shvet1]

As per my experience with Aspen+ this software gives a low quality of modelling tees and 2-phase flow. As Aspen+ is strongly interconnected to Hysys I believe the situation is the same. My opinion is this utility was developed for a preliminary only prediction of hydraulic of a simple pipe segments like a straight long-distance piping.

 

Note that Aspen prefers not to describe in details how utilities actually work. Results are user's responsibility only.

 

E.g. Pipiphase gives more detailed and therefore reliable reports.

Edited by shvet1, 20 June 2023 - 03:38 AM.

[Bobby Strain]

After a bit more thought I must revise my advice a bit. Since all the correlations are based on correcting the total pressure loss by subtracting the static head for vertical upflow, one must do the same when using the correlations.

 

Bobby

[tdeb]

Dear all, 

 

Thanks for your answers, and apologize my delayed reply. To clarify, I should emphasize that I'm aware of the fact that static pressure drop is independent of flow rate, and would be constant for any system if/only if there is no evaporation in line. When there's evaporation in line, which is my case, there will be some dependence of the static pressure drop with the flow rate, as more flow rate means more friction pressure drop, hence more evaporation, hence less static pressure drop. Nevertheless, my question was about how much evaporation affects static pressure drop. As I said, if the system is liquid only, the static pressure drop in my system (with a height difference of 40 m) is about 2 kg/cm2g. But, my system has a massic vapour fraction of 5% in the heighest point, and Hysys is estimating a static pressure drop of around 0,2 kg/cm2. This is what I find weird, because I was not expecting such a big effect on static pressure drop with such low evaporation

Edited by tdeb, 23 June 2023 - 05:39 AM.

[Bobby Strain]

All the correlations evaluate static head based on insitu density. Which is, indeed, a function of flow rate.

 

Bobby