2 replies to this topic

[Elochi]

I am experiencing difficulty modeling a process using ASPEN HYSYS 2006. I wish to design a distillation column that maximizes the yield of kerosene from diesel feed. So far, I have only been able to install the oil in the flowsheet and define the feed. The kerosene has a distillation range within the diesel and so I was thinking to use a column with a side-stripper where the light ends will come off as the distillate, the kerosene as the side product, and diesel residual as the bottoms. The Hysys tutorials online have been helpful but converging the column (after specifying recoveries and stream compositions) has been the road block which has seriously held up my progress.

I will be grateful to anyone that can offer assistance on this matter. Thanks in anticipation!

[Milutin]

Hi Elochi,

Looking at feed properties you have already stabilized ULSD, no light end presented. If you want to satisfy kero distillation specification, you choose right configuration.
Essentially it is atmospheric distillation column, with naphtha as overhead, kero side stream and diesel as bottom. You can look at Hysys example for crude atmospheric distillation column, and adjust it to your example.

To really maximize kero production you should not stick to distillation 10% specification but only to flash point.

Regards,

Milutin

[Zauberberg]

Elochi,

From what I can see, you need some kind of "diesel splitter" column. As Milutin pointed out, since you don't have any light ends in your feed, two-product tower would be just fine.

1. Be sure that your fluid characterization parameters are OK. Matching bulk density is important, but I wouldn't rely on ASTM D86 and vapor pressure predictions by Hysys. Transport properties are also week point in Aspen software. You can tune your pseudocomponent breakdown pattern in order to reach desired values of bulk properties.

2. Regarding column convergence, try to relax your model by using one of these approaches:
- Specify ranges of convergence parameters instead of fixed values
- Don't enter physical properties for column specs
- Don't enter mass or any other flow of bottoms product as a spec. Try to specify temperature profile, condenser temperature etc.
- Be sure that pressure drop across the column is within expected ranges: 0.005-0.01 bar per tray

I guess you are using Peng-Robinson thermo package for the column model. I think this is the best model for non-polar hydrocarbon systems, with lot of experience and experimental data incorporated.

Best of luck,