3 replies to this topic

[neriet]

I have been suggested to select CPA Cubic Plus Association for a water-methanol and water-ethanol separations (water, alcohols plus several other components),
my copy of Prode Properties software (Base version) includes three CPA packages, PRXCPA(VDW) , PRXCPA-NRTL(P-HV) and PRXCPA-NRTL(P-LVM),
are these CPA equivalent ?
Also, my version of Prode doesn't include the CPA package for enthalpy, which package do you recommend (for water-alcohols) ?
Thanks for help.

[PaoloPemi]

see also this thread,

https://www.cheresou...ation-in-prode/

 

Prode Properties Base has 4 packages based on CPA (one derived from Soave-RK and three from Peng Robinson),

two with VdW mixing rules and two with complex (modified Huron Vidal and modified LCVM) mixing rules,

the models with complex mixing rules are more flexible (i.e. NRTL with temperature dependent parameters) but a bit slower,

I would prefer PRXCPA-NRTL because of additional flexibility ,  P-HV means modified Huron Vidal while P-LCVM the Linear Combination of Vidal and Michelsen rules,

for enthalpy you can select the base model with complex mixing rules (i.e. PRX-NRTL (P-HV) or (P-LCVM)

another possibility would the MBWR or similar models which however are available in extended versions, only.

[neriet]

thanks,

in your opinion, do the existing BIPs provide good accuracy or should I run the VLE/LLE data-regression utility ?

[PaoloPemi]

that depends from operating conditions (in your case),

you can inspect the range of values in data regression, these values are stored (with average error) in Prode Properties, BIPs database

in this case (water-methanol) error is low even if NRTL doesn't include temperature dependences (U12-T and U21-T are 0),

the association terms in CPA provide the required interaction parameters,

 

anyway, if you have access to VLE data points you can calculate the BIPs (with the included VLE/LLE/SLE data regression utility)

Edited by PaoloPemi, 16 January 2018 - 12:04 PM.