1 reply to this topic

[jbmenius]

Hi All,
I am simulating the separation of C2-C4 alkenes and alkanes from their corresponding alcohols and water. Although the components have very different boiling points and separate easily at atmospheric using only a flash drum, my stream is at 650 psi (44 atm). According to Aspen, at these conditions I must use a distillation column to get the same separation as a flash drum at ambient. However, the reboiler operates at 220 C. As you can see from the little table in the excel attachment, this is well above the Tc of several components. The low-Tc alkanes are the largest molar components, so the average Tc is only 95 C. The top of the column is at about 85 C.

My concern is that the mixture would actually be supercritical at these conditions, making a vapor-liquid based separation impossible. Obviously, the concentration of alcohols and water is greatest at the bottom of the column where the temp is highest, and these have higher Tc, so maybe that keeps things from going above the critical point. Likewise, the lowest Tc components are most concentrated at the cooler end of the column. Also, the pressure is above the Pc of many of the components.

I am using the SR-Polar property method for all simulations. Aspen says its results are alright above critical.

Does anyone know if this separation is realistic, or just extrapolated results? Thank you for your help.

 Critical Points.xlsx   9.4KB   29 downloads

[abhi_agrawa]

Jmenius,

The average Tc reported in you spreadsheet is the pseudo-critical temperature and not the true Tc. The true critical temperature should be estimated from the phase envelope. In case you are really above the true critical temperature and pressure, suggest that you reduce the pressure of the column. This will also reduce the temperature at the bottom of the column.

A very important aspect of simulation is choosing the correct thermodynamic model. Ideally, it would be great if you have the VLE data at the operating conditions and then you can check this data against the predictions from the simulator. In reality, this is not the case, often we do not have data at the operating conditions. Suggest you try to obtain VLE data of several of your components and check the predictions from ASPEN against that. This exercise would give you some confidence on the model.

Hope this helps,
Abhishek