5 replies to this topic

[marya1997]

Hello everyone,

I need to simulate the attached flowsheet in HYSYS, but I don't know exactly where to start. Should I simulate and run the distillation column with the condenser and reboiler, and then hold the solver, remove the condenser, and substitute my desired equipment? Or should I use another approach? I would be very appreciative of your help and advice, or if you could introduce me to some tutorials, videos, books, etc. to guide me.

Thank you

[marya1997]

i think my file in is not attached. here is my flowsheet

Attached Files

•  Untitled5.png   425.79KB   1 downloads

[Pilesar]

Use fewer theoretical stages for the column. Return the sidestripper vapor to the stage just above the sidestripper feed. Solve the crude column simply first, then add complexity. Figure out the approximate flow of side products from analyzing your feed stream before you begin. For first solution, leave out the sidestrippers and set the side draw rates. Spec the naphtha 95% temp and the overflash percentage (the extra amount of vaporization in addition to the feed - about 3%.)  Vary the top and bottom duties. After you get solution, add in the sidestrippers and vary their respective draw rates to meet the draw stream 95% specs. Crude columns are complex so get the simplified convergences first. Grayson Streed is a good thermo package to use. Your top product will probably not match specs because light ends are difficult to get accurately. The other product streams are more important to get right. I just searched internet for 'crude column simulation' and got 12 million hits. One of those references surely goes into more detail than what I gave here.

[Pilesar]

Very important is how you characterize the feed stream. You must use enough pseudo cuts to have overlap between side products. Feed characterization is critical to your model so give it the attention it deserves.

[Pilesar]

You are not modelling a reaction in the column so there must be a material balance: components in = components out. One of the secrets to good crude column modelling is to first find the composition and flowrates of the product streams if possible from plant measurements and analyses. Since that is what you are trying to match in the simulation, define the product streams into your model then combine them in a mixer unit to form the feed to the column. The product stream compositions are much more precisely measured. The feed assay is likely not very current and will not generate the measured product streams so don't even use the feed assay if you have better measurements of the product streams. Then adjust your column parameters to try to get the simulation results close to the measured product streams. It may sound like a cheat, but it is the best way to approach the answers you believe correct from plant measurements. 

[Pilesar]

Bobby Strain just posted a link containing an example 'how to' for a crude column simulation using Hysys which might be useful. Here is the link: https://people.mines...n-HYSYS_v10.pdf