Multicomponent Distillation

[_Greg_]

Hi folks

I'm currently in my last year of my chemical engineering degree and doing a design project on the manufacture of acrylic acid. I have the job of designing a solvent recovery column, where a mixture of Methylisobutylketone (MIBK) solvent (heavy key) and acrylic acid (light key) are to be separated.

To be honest with you guys I'v not had many good grades throughout my course and it's at the design stages where I suffer as a result, and I'v been stressing over this a lot, time is just passing me by and don't seem to be getting anywhere, so whould really appreciate some help.

So firstly I'll show you my feed composition:

Component-----------------Feed (kg/s) 60*C, 13kPa
Propene-----------------------0
Oxygen------------------------0
Nitrogen-----------------------0.0251
Carbon Dioxide-----------------0.0500
Water-------------------------0.5391
Acrylic Acid--------------------8.5548 (LK)
Acrolein-----------------------7.5365
Acetic Acid--------------------0.2035
MIBK--------------------------29.3277 (HK)

Total---------------------------46.2367

My Assumptions are:

· Key Components: MIBK and Acrylic Acid
· Pseudo-Binary mixture
· 100% recovery of Acrylic & Acetic Acid in bottoms
· 0% solvent & water in bottoms
· From the research I'v done, packing seems to be the way to go for this purpose.

Here's a diagram of my section:

So overhead we should have water and MIBK going to a partial condenser, the liquid phase goes to a gravity separator where MIBK and water separates while the remaining vapour (majority water) goes to a 2nd condenser. MIBK is refluxed to the column from the gravity separator while the water from the separator and 2nd consenser mixes and goes to a raffinate stripper for treatment.

This is a design we agreed on as a group, and was found in a library journal, though I'm slightly confused as to why there should be 2 condensers, why not just condense it all and separate the whole lot?

Anyway my question, I don't know where to start with this design! where do I start?

I'v seen multicomponent distillation examples with light hydrocarbons, propane, butane etc but nothing like the mixture I have. Also it's not exactly a straight forward reflux setup, and have no idea how to model this on Aspen HYSYS.

So basically can anyone help me on where to start? Maybe list the sequence of calculations I should do?

I'v tried a shortcut design on HYSYS, but it doesn't include this reflux setup. Also it only works if I set the liquid fraction of the top product stream to 1, which is confusing, surely 0?

I have done reading through Coulson and Richardson vol 6, but again, I'm not the brightest and still really confused. I think I need to get a proper simulation working, get those K values (vle coefficients) and somehow do a manual calulation (can't rely on HYSYS alone)

If you managed to read this far, I thank you! Not a very clear cut question

 

[_Greg_]

Please, anyone, I seriously feel like dropping out at this rate :(

 

[piuz98]

I don't know how the program that you use in run this simulation, but as I know you need to set up the output of the simulation to run the simulation. and from your diagram, I think you should not put two condenser for line 23 & 24. and also the line 29 will give you some error as well when you run the simulation.

 

[Vulgar]

Hi

WHat do you mean by, you don't know how to model this on Aspen? Does Aspen not have all components?

I think it comes down to the thermodynamic models you chose for the simulation, and any components you don't have, you can try and let Aspen figure out the thermodynamic properties with packages like UNIFAQ or UNIQUAC.

And for the simulation itself, I think Aspen will tell you when it can simulate, you just need to input all the data so that the program has no degrees of freedom left.

 

[alphy]

Hi there, it seems like you are working on a challenging project. Multicomponent distillation can be a complex process, especially when separating components with similar boiling points, like MIBK and acrylic acid. It's great that you have identified the key components and made some assumptions to simplify your design.

To start, I would suggest doing some preliminary calculations to determine the number of theoretical stages and reflux ratio needed for your column. This can be done using the Fenske-Underwood-Gilliland (FUG) method or the McCabe-Thiele method. These calculations will give you an idea of the column's design and help guide your decisions on packing, reflux setup, and other design parameters.

Regarding the use of two condensers, it is likely due to the high water content in your feed. Water is a polar solvent and can form azeotropes with many solvents, including MIBK. The use of two condensers allows for better separation of the water and MIBK, resulting in a purer MIBK stream that can be refluxed back to the column.

As for modeling this process on Aspen HYSYS, I would suggest starting with a simple two-component system (MIBK and acrylic acid) and adjusting the parameters to match your feed composition. You can then add in the other components and make adjustments accordingly. It may also be helpful to consult with your professors or other experts in the field for guidance on using the software.

In terms of your design, it's important to consider the energy and material balances, as well as the economics of your process. You may also want to consider the use of additional equipment, such as a decanter or heat exchanger, to optimize the separation.

Overall, the key to a successful design is to thoroughly understand the process and make informed decisions based on calculations and simulations. It may take some time and effort, but don't be discouraged. With persistence and the help of your colleagues and mentors, you can successfully design a multicomponent distillation column for the separation of MIBK and acrylic acid. Good luck!