6 replies to this topic

[Said Salim]

Dear All,

Whenever I am selecting the Peng-Robinson fluid package in Hysys, I find estimated values of Binary Interaction Parameters (BIPs) that are used for the calculation. Anyone have an idea by what method those BIPs are estimated ? is it Group Contribution method? or another method? or they are determined experimentally?

[serra]

(as far as I know) the extended PR available in Aspen should include Bips calculated (mainly) from experimental data,

some Peng Robinson variants allow to specify temperature dependent BIPs (2 or more values, for example the different Extended Peng Robinson versions included in Prode allow to specify up to 4 values),

usually these (temperature dependent) BIps are calculated from experimental VLE data (see DDB, GPSA or similar sources),

most simulators include a data regression procedure and with that you can calculated BIPs from VLE / LLE data or from values estimated with Unifac etc.

[MrShorty]

According to this document http://www.diquima.u...elsV732-Ref.pdf (see page 49)

The Aspen Physical Property System also has built-in kij and lij for a large number of component pairs in the EOS-LIT databank from Knapp et al. These parameters are used automatically with the PENG-ROB property method.

Knapp et al appears to be the reference for the DeChema Databank. It doesn't say what it does for binary pairs that are not in the literature database.

Edited by MrShorty, 03 May 2017 - 01:40 PM.

[Said Salim]

What about the radio bottun HC-HC estimate in HYSYS? Is it indicating the calculation method?

[serra]

There are methods for estimating BIPs of components not in std. library (for example pseudo components) , see API TDB etc. for details, probably (not sure about exact method included) Aspen does that,

personally (where possible) I prefer to calculate  BIPs from VLE / LLE etc. data points (I have Prode but the same for Aspen)

these procedures allow to compare the errors for the different models and select the most suitable for that specific application... 

Edited by serra, 04 May 2017 - 10:08 AM.

[MrShorty]

I do not find any description of the algorithms behind those estimates. This document seems to be the most common that comes up http://people.clarks...SpropSelect.pdfand it simply states that

For non-library or hydrocarbon pseudo components, HC-HC interaction parameters will be generated automatically by HYSYS for improved VLE property predictions. (see page A9 and A15)

 

This document https://www.slidesha...anfuels20065usr(section 3.2.1) suggests that these methods of estimating PR bip's may be proprietary, so there may not be full documentation or descriptions available for exactly how Hysys performs these estimates.

[DanW]

It probably uses the following method, as DWSIM does.