4 replies to this topic

[Omani]

Hi,
I am working on a design of an extractive distillation column of benzene (light key) from a mixture of-mainly- xylene isomers. Hysys gives me a column with 75 trays (100% efficiency) and a reflux ratio of 100.
My questions are:
1) Is there a better solvent to use instead of the one I chose? (I am using benzonitrile)?
2) What is the practical maximum acceptable height of a distillation column (weight, stress .etc)?
3)What "good" assumptions can I make when deciding to spend on OPEX or CAPEX based on 2 above?
4) What do I need to look for if |Idecide to split up the column into two sections (to reduce height)?

My other question is regarding diameter of the column. I need to use densities of liquid and vapur for that, but these densities will keep changing as a function of column height. Should I take the averages?



Thanks a lot for your help in advance
Omani

[Omani]

Hi,
I am working on a design of an extractive distillation column of benzene (light key) from a mixture of-mainly- xylene isomers. Hysys gives me a column with 75 trays (100% efficiency) and a reflux ratio of 100.
My questions are:
1) Is there a better solvent to use instead of the one I chose? (I am using benzonitrile)?
2) What is the practical maximum acceptable height of a distillation column (weight, stress .etc)?
3)What "good" assumptions can I make when deciding to spend on OPEX or CAPEX based on 2 above?
4) What do I need to look for if |Idecide to split up the column into two sections (to reduce height)?

My other question is regarding diameter of the column. I need to use densities of liquid and vapur for that, but these densities will keep changing as a function of column height. Should I take the averages?



Thanks a lot for your help in advance
Omani

Sorry,
The light key is Ethyl Benzene (not Benzene) Thanks

[abhi_agrawa]

Omani,

The maximum height of a column is decided by two factors:

1) Local aviation norms: depending upon these I have seen columns being restricted. These depend on how far are you from an airport. To know the same you will have to get in touch with the civil aviation authorities.

2) The length to diameter ration of the column. This is usually restricted to 20 - 25. This is a mechanical limitation.

Usually splitting the column is done such that both the columns are similar in height. This ensures that you do not have a set where one of the columns if of the maximum allowed height and the other is quite small. You have to understand, the in most of the cases the columns have monkey ladder for climbing up. So to go to the top of the tower will take a huge effort and the operators in the field may not be willing to put in that effort.

To decide between OPEX and CAPEX (which I assume stand for OPerating EXpenditure and CApital EXpenditure), you will have to do a return of investment analysis and see how much time does it take to break even.

The column is usually divided into several zones depending on the vapor/liquid traffic, draws, feeds etc. Then each zone is sized on the the maximum traffic in that zone. Do not size by average properties because then there will be some trays which will see a higher traffic as compared to the average and there flooding will be initiated. This will eventually flood the entire column. Therefore the sizing is always done on the basis of maximum loads.

Hope this helps,
abhishek

[Omani]

Thanks a lot

The other thing that I am struggling with is to try and find out if there is any company that carries out ethylebenzene separation from mixed xylenes using extractive distillation. I have come across many patents which claim that it it feasible. However, I have tried all the solvents they "tested" in labs, but none was economically feasible according to the separation factor (S)<10. I used hysys to simulate using extended NRTL fluid package).
The unit Ops that I have seen in industry is SMB (simulated moving bed) adsorption, but it is kinda complicated giving the knowledge and the resources (and time) that I have for my desig project.#



Thanks

[abhi_agrawa]

Omani

I have no idea about Ethylbenzene from mixed Xylene, so can't help you there. But I'd like to caution you on the use of simulators. A simulator is nothing more than a tool to do calculations for you and it gives you results based on the equations entered in it. So when you use a simulator for a process, about which you do not know much then the way out is get all the possible VLE/LLE/VLLE data and then try to match them from the predictions of the simulator. You can have a look at DECHEMA/Journal of Chem. Eng. and Data etc. for this information. This is a very important step. While doing this you may notice that some of the thermodynamic parameters in te simulator may need to be modified. Only when you have been able to get a reasonable prediction of this, you can use the thermodynamic method you have chosen. Else you can not be sure of the predictions.

I'll give you a small example illustrating the above: in ASPEN the interaction parameter between H2S and CH4 is missing. Therefore if you use ASPEN and then HYSYS for the separation from a mixture that contains these components, your predictions will be different. So you may be led to believe that HYSYS is better for such applications. However, if you enter the correct interaction parameter in ASPEN, you will get same results. Therefore knowing the correct interaction parameters is very important.

Hope this helps,
Abhishek