38 replies to this topic

[staffel]

GPA has published some data,

KDB has some data points,

"http://www.cheric.or...926&cno2=1"

which you can easily regress with PRODE PROPERTIES,
but I do not know how accuarate these VLE points are.

Edited by staffel, 26 November 2013 - 06:32 AM.

[mrbabu]

thanks Staffel, that site includes indeed many VLE data points which would result useful for me, do you have an idea if they are experimental or predicted ? I am unable to get a reply from the web master.

[PaoloPemi]

mrbabu,

it is my feeling that many points have been calculated (not measured)

but I have no data to support this affirmation (you should contact the web master)

[mrbabu]

thanks Paolo,

I see that PRODE databank stores a estimate of errors when regressing BIPs,

do you think it's correct to select the most suitable thermo package comparing those errors ?

[PaoloPemi]

yes, you can select the thermo package on the basis of best VLE data fitting as reported in PRODE PROPERTIES,

those values are related to the accuracy  of each thermo package to fit a specific VLE data set.

[staffel]

you may consider the range of temperatures and pressures reported per each BIP set, too,

that should give a clear indication (see the dialog)

 

Attached Files

•  BIP_PRODE.jpg   54.13KB   1 downloads

[frpe]

thanks Staffel,

does the range of application for those BIPs be limited to that range ?

[chemdoc]

I would expect that models as EOS (Peng Robinson, Soave etc.)

should allow to utilize temperature dependent BIPs outside the specified range,

(for example I get accurate results for natural gas + water at high pressures)

while models based on liquid activity (Winson, Uniquac etc.)

should be a bit more limited.

[mrbabu]

thanks,

I understand that I can select the thermo package comparing the reported errors

in data fitting as shown in Prode Data Bank.

[frpe]

Yes,
that is my understanding, too.

[serra]

as said by others accurate BIPs are required to calculate VLE,LLE,SLE
in some cases the fluid  may behave as ideal or the EOS can fit VLE points without corrections but in many cases to model accurately VLE/LLE at high pressures (for example in phase envelope) you need BIPs,
simulators as Prode include a large databank, fitting experimental points is another option as discussed in previous posts and there are several commercial and public collections of VLE/LLE equilibrium data, data regression is easy,
I agree that X-Y data fitting error available in Prode for each BIP data set gives a good indication to select the model.

[bponcet]

can I include the hydrate formation temperature curve available in Prode Properties

in phase envelope diagram calculated by Prode ?

How can I include this curve ?

[marchem]

yes, and it's very easy with VBA,

with Prode Properties you can calculate equilibrium points given pressure or temperature,

have you examined the VBA code in hydrate.xls (included in Prode Properties distribution) ?

 

For calculating equilibrium points given temperature

 

    nrpt = 10
    Tmin = 250
    Tmax = 320
    dT = (Tmax - Tmin) / nrpt
    For i = 0 To nrpt - 1
        Tcalc = UMCS(Tmin + i * dT, UM, getUMC(UM))
        Pcalc = HPFORM(stream, Tcalc, model)
        Cells(12 + i, 12) = Tcalc
        Cells(12 + i, 13) = Pcalc
     Next i

 

 

the same for calculating equilibrium points given pressure,

in that case utilize the HTFORM() macro available in Excel.

 

Of course you can put macros HTFORM() or HPFORM() in a cell

and get directly the values

[mrbabu]

thanks marchem,

indeed it's useful to put all the lines (vapor, liquid, hydrate) in a single graph