Try the Cubic+ Association method (CPA). This will give you a binary azeotrope. Whether it is close or correct in its prediction is another question.

Is the reaction taking place in the vapor phase? If not, Aspen does not allow partial pressure as a driving force for liquid phase reactions.

What is the reaction you're trying to include?

Brian Hanley Thank you for your help! This method definitely helped, but the azeotrope is at a lower molar fraction of CH3OH than expected, so I will look into that!

Yes, it takes place in the vapor phase. The reaction I'm working on has a rate equation depending on several constants and the partial pressures. However, I don't know how to incorporate such as specific type of reaction kinetics. I know how to incorporate power law, LHHW etc., but that is not suitable for this reaction.

You can regress parameters for any VLE model using experimental data. By default (I think) Aspen gives access to the NIST database of experimental data. There is a good amount of VLE data there for your system. In theory, then, you can update the default model parameters for any VLE method so as to give a "best fit" to the data sets you choose to include. That being said, regressing parameters for VLE models is not so straightforward.

Further, if the reaction kinetics you have don't conform to any of the built-in reaction models, you can always write and compile your own FORTRAN subroutine (described in the attached manual). Once again, this is a fairly difficult thing to do if you've never done it before.

Thank you for your extensive explaination and I will look into the manual!

problem 1: You should chose all databases for the NRTL model.

problem 2: I suggest that you should see thee help file from the Aspen Plus.

I hope these answers will help you.