Excel library for process calc's including distillation

[cmarc]

thanks Carlo,
they are really good for my actual project

[tmander]

can someone explain the different options for phase envelope calc's in Properties ? I am familiar with the minimum Gibbs Energy criteria which is the standard for process simulators but I do not understand validate EOS roots according state.

[carlo.stenali]

tmander, my copy has three different options

1) standard
2) select EOS root for minimum Gibbs Energy
3) Validate EOS root according state

the usual way is to select the EOS root for minimum Gibbs energy, validating EOS root introduces an additional test which can result useful in some cases, see for example the mixture CH4 0.85 H2S 0.15 (SRK-API) , there are more complex examples with Wong Sandler or Huron Vidal mixing rules.

[tmander]

thanks Carlo,
a interesting example, if I understand the point with that option the procedure attempts to validate roots dicharging the points with abnormal density.

[carlo.stenali]

yes, that is the way.
Consider the common test case mixture,
CH4 0.7 H2S 0.15 CO2 0.15
at 200 K and 2000 Bar
SRK gives only one root Z = 4.194
is that the value of a liquid or a gas root ? Really a unusual value.
If in Properties you set on Models dialog the Detect Single Phase State to "From Gibbs Energy"
it returns (correctly) a liquid state
the same at 220 K and 2000 Bar
however if you do a calc at 240 K and 2000 Bar with "Gibbs Energy" option the procedure returns vapor state.
Now if you set on Models dialog the Detect Single Phase State to "From Isothermal Compressibilty" the procedure returns liquid state at 240 K and 2000 Bar
What I would say is that there are many options and you should spend some time to understand how they work.

[tmander]

thanks Carlo,
that (detecting the correct single phase) is a quite common problem for all simulators, with EOS I think the selection criteria based on isothermal compressibility is the best in therms of efficincy, some software do calculate a bubble (or dew) line but it is a slow procedure (when doing a lot of t,p flash operations)

[cmarc]

I am calculating 2-Butanol Water VLE at 10 atm in Properties, I suspect regular (for gas fugacity) is not good at that pressure, if I wish to test Peng Robinson (with Wong Sandler mixing rules) should I use the BIPs utility to regress the VLE data (see the previous post by Carlo Stenali) for the specific model adopted ?

[carlo.stenali]

cmarc,
yes, with Wong Sandler you should regress BIPs (you may use the data regression utility in Properties, see "http://www.prode.com/docs/pppman.pdf" for information), there are three (Wilson / UNIQUAC) or four (NRTL) parameters to calculate, the additional parameter being kij , note that the data regression utility can be very slow in some cases, with NRTL (four parmeters to calculate), it takes about 10 seconds in my computer, with complex mixing rules I prefer to set the extended Alpha function (which is the default) as this should reduce errors in vapor pressure correlation (which is important in VLE)

[cmarc]

Carlo,
does Properties use as default for Soave Redlich Kwong and Peng Robinson the TWU Alpha function ? I see in databank the constants A,B,C but these are quite diferent from the values which I got from a different simulator, if I wish to compare VLE results should I set the SRK, PR Alpha functions ?

[carlo.stenali]

with natural gas mixtures I prefer base alpha function, phase envelopes seem more "realistic" and I am not sure if extended Alpha function does produce reliable results above critical points, see for example a simple methane-n-butane phase diagram calculated with PR and base Alpha function or extended (TWU) Alpha function, I get quite different values
with base
PC 130 Bar TC 326K CricoT 356K CricoP 133 Bar
with extended
PC 138 Bar TC 323K CricoT 355K CricoP 140 Bar

said this I must add that with glycole, water, methanol etc. the extended Alpha function works better so it depends from your application.

[cmarc]

thanks,
which thermodynamics should I define for a glycol dehydration absorber ?
GPSA chapter 20 has VLE data for TEG and others, what about PR with Wong Sandler mixing rules ?

[carlo.stenali]

for TEG (triethylene glycol) and water VLE the PR-WS works well, for enthalpy PR do a decent result, you should set the extended Alpha function as base alpha doesn't fit well these components (other components, usually hydrocarbons, work well).
In the past I did use PR with specific BIPs getting quite good results (at least at low pressures), however WS or HV mixing rules should do a much better work, you must spend some time for tuning parameters as these mixing rules require at least the Kij