2 replies to this topic

[pdender]

In version 1.2 PRODE PROPERTIES BASE includes the 5 parameters formulations for Peng Robinson and Soave Redlich Kwong, in addition there are 7 parameters formulations for CPA (both PR and SRK variants),

previously the extended 5 and 7 parameters EOS formulations were available only in separate versions. 

Which are the main advtantages of these formulations ? 

I have found that densities and enthalpies are more accurate (I have a limited number of experimental data to compare) compared with 3 parameter formulations but I think there are advantages also for VLE /LLE , 

is that true for all components or some specific chemical ?

For example with hydrocarbons should I use Peng Robinson standard or extended ?

[marchem]

according my experience (the 5 parameters formulation in PRODE has been available since version 1.7) there are some advantages and some problems,

for polar fluids as water, glycols etc. it's very good for both phase equilibria and properties as enthalpies, densities etc.

For example you can expect a better accuracy solving a column (VLE / LLE + H), a separation, or some other operations where VLE, enthalpy, entropy and volumes are not well predicted by Std. or extended versions of Peng Robinson / Soave Redlich Kwong

The same for some "supercritical" fluids as hydrogen.

The problems are mainly related to some "strange" phase envelopes which may be generated with 5 parameters formulation (for phase envelopes I prefer base formulation).

Finally, I do not utilize frequently the 7 parameters CPA, mainly for phase equilibria when components as water and glycols etc. are present in the mixture,
it's a bit slow compared to std. EOS (without association term) but should produce very good results
(difficult to verify without access to a laboratory to validate data)

Edited by marchem, 31 July 2013 - 03:51 AM.

[PaoloPemi]

I some cases the "strange" phase diagrams may be originated by abnormal values for BIPs,

PRODE includes a relatiively large number of BIPs (more than 30000) but several (for example hydrocarbons-water which have usually values of 0.5 for Soave Redlich Kwong and Peng Robinson)

are from literature not calculated by software (data-regression from VLE experimental points) taking in account the dfferent alpha functions.

In these cases one can utilize the data-regression utility to calculate new values.