21 replies to this topic

[ChemiFreak]

Dear All,

 

Could u please explain about the integral method to be used in PSV sizing calculation when the Compressibility Factor is not within the range of 0.8 and 1.1?

[breizh]

hi,

Consider this document to support your query.

Hope this helps

 

Breizh

[serra]

see API 520, the procedures discussed in API and other sources attempt to locate vena contracta conditions,

CHERESOURCES has a Excel page to size a PSV with different methods and comparing results

http://www.cheresour...ng-temperature/

as alternative you can calculate critical mass flux estimating speed of sound with rigorous methods based for example on EOS, this method allows a quick (and more accurate) solution,  

http://prodesoftware...ons.blogspot.de

Edited by serra, 08 July 2014 - 07:47 AM.

[latexman]

Take a look at this FAQ I wrote over at Eng-Tips.  I think it's exactly what you asked for.

 

http://www.eng-tips....qs.cfm?fid=1293

[ChemiFreak]

Thank u all for ur valuable information.

[ChemiFreak]

Latexman,

 

Firstly thanks for the source provided. Could u please elaborate on how Gn is calculated. Is it by calculating the velocity or volume???

[latexman]

Read down into the FAQ until you get to:

 

"G_n = [ρ]_n ( - 2g_c [∫] dP/[ρ] )^1/2

Evaluate [∫] dP/[ρ] (the RHS of Equation 1) numerically from P_o to P_n until G_n reaches a maximum (sonic flow) OR P_n = P_BP (subsonic flow)"

 

That's the key! (I hope you can interpret the TGML code from my copy paste of the FAQ.)

 

Then look at the spreadsheet output.  The VBA in the spreadsheet starts at Po and To and calculates [&rho]o.  Then in the next row down, it increments the pressure down by dP (usually 1 psi) for Pn.  It calculates Tn isentropically due to pressure reduction.  It calculates [&rho]n using the ideal gas law.  Any EOS can be used.  Then, it calculates and sums ([&delta]P/[&rho]ave) in each pressure increment.  Then it calculates G for each increment using G_n = [ρ]_n ( - 2g_c [∫] dP/[ρ] )^1/2.

Edited by latexman, 10 July 2014 - 02:56 PM.

[breizh]

Hi,

Some additional input about the method.

Breizh

[ChemiFreak]

Thanks Latexman and Breizh.

 

Latexman,

 

Am not able to download the Spreadsheet. Could u post the same link in Chereources Forum if possible.

 

Thanks in advance.

[serra]

see post #21 in thread
http://www.cheresour...ng-temperature/

there is the Excel page which allows to calculate a PSV with

1) integration method (supercritical and two phase flow)
2) API formulation with real cp/cv
3) API formulation with ideal cp/cv
4) hydrocarbon's magazine paper

the psvcompare1.xls page allows rigorous design/rate of a PSV comparing different methods

the procedure calls Prode Properties for fluid thermophysical properties
and flash operations, a free copy of Prode is available here

http://www.prode.com/en/properties.htm

the Excel example psvcompare1.xls works fine in my tests
and you can look at VBA code before to start writing a new one.

Finally, if you wish to solve directly the nozzle model
thake a look at ISPF() method in Prode which adopts a
direct method.

Edited by serra, 11 July 2014 - 01:16 AM.

[ChemiFreak]

Thanks Serra.

 

I did download the excel that you have mentioned. But am not that good in VBA to proceed further.

[serra]

for solving the integration method you need some knowledge about programming

as you need to code a iterative procedure,

 

as alternative (to a iterative procedure) you can solve much more flash operations

for example in Excel with a process library or a simulator and

interpolate results for finding vena contracta conditions,

 

there are no simpler alternatives.

[ChemiFreak]

Thanks Serra.

 

Ill start learning VBA.

[latexman]

ChemiFreak,

 

That was just a copy/paste of my spreadsheet into the FAQ text editor; not an active spreadsheet or link.  I wrote it to test and understand the method.  I haven't opened it in 7 years.  It's not user friendly and after re-looking at it I see some issues.

 

One of the ways to learn VBA quickly is to record a macro and then open up the macro and study it.

[ChemiFreak]

Thanks Latexman. 

 

Will definitely take your advice.

[aqassam]

FYI: Aspen HYSYS v8.3+ has the direct integration method available in the Safety Analysis environment. You can just do your analysis in HYSYS. 

[frpe]

in Excel or with VBA you may call the macro ISPF() available in Prode Properties to test your results,

ISPF() models a PSV with

HEM (Homogeneous Equilibrium Model)

HNE (Homogeneous Non-Equilibrium Model)

NHNE (Non-Homogeneous Non-Equilibrium Model)

the solution is faster that direct integration being based on a direct approach (evaluation of critical condition),

it is more accurate, too.

[ChemiFreak]

Thanks aqassam and frpe for the info.

 

Aqassam,

 

I did try it in UNISIM R430 using databook. But I find the results contradicting with the numerical method (stated by Latexman) where temperature and density are directly proportional but in UNISIM they are inversely proportional. Could u help me on this???

[frpe]

I fear you are confusing properties,
V=(ZRT)/P
volume is proportional to temperature
density is the inverse of volume or D = 1/V

if you do not have a good knowledge of thermodynamics
better to use a well tested code such as PRODE PROPERTIES,

differently, the Excel page suggested by serra

http://www.cheresour...ng-temperature/

has exactly the code which you need,
but you must learn VBA programming as suggested by others.

[ChemiFreak]

Thanks frpe for your advice.

 

Your post seem to deviate from our current discussion. I just wish to know if someone could help me to solve the direct integration method using UNISIM (having some issues as explained in my previous post) since am not able to access HYSYS, as suggested by our fellow colleague, as of now. Wish u could help me to solve this.

[frpe]

have you compared your results with the Excel page suggested by serra ?

That would be a simple way to learn the method.

[ChemiFreak]

Thanks frpe.

 

I'll do so and get back to you.