8 replies to this topic

[engeng]

Hello for All

Please I would like to know, in general, how many trays should I need (instead of packing) in design of topped crude vacuum distillation, in case I want to separate LVGO,  HVGO, Slop oil, and asphalt, i. e. the number of trays for each section.

 

Thank You in advance

 

[Technical Bard]

I would review the texts by Gerald Kaes (Refinery Process Simulation) and Henry Kister (Distillation Design) to get this answer.

[engeng]

Thank you for your reply, in the text by Gerald Kaes, it is referred to theoretical trays in the simulation of vacuum tower, my question now is : it is true to use theoretical trays alone without efficiency, i. e. no. of theoretical trays = no. of actual trays?

[PingPong]

No, number of actual trays is always higher than number of theoretical trays.

That is especially true in a topped crude vacuum distillation column due to high liquid viscosity and low liquid residence time on the trays.

 

I expect that above the feed you would need at least 3 actual trays per theoretical stage.

Below the feed in the stripping section use at least 5 actual trays per theoretical stage. The latter depends very much on the type of trays you would use in the stripping section. If simple baffle trays then use 10 actual per theoretical.

 

Nowadays nobody would design topped crude vacuum column using trays above the feed, so I strongly suggest you don't either.

[engeng]

Thank You PingPong for your clarifications. Please with simulation of vacuum tower or others, the number of actual trays should be known because the process simulator, for example Hysys need it and then it converts to height of packing, so how many trays I should enter for HVGO section and LVGO ?

[PingPong]

I don't how what your vacuum column is going to look like, what product specs you put on LVGO and HVGO.

If there is no fractionation between LVGO and HVGO, which is usually the case, then you can use 2 theoretical stages for the LVGO pumparound, 2 theoretical stages for the HVGO pumparound, 2 or 3 theoretical stages for the wash section (between HVGO draw and column feed) and 1 or 2 theoretical stages for the stripping section (below column feed).

 

If you want to improve (for whatever reason) the separation between LVGO and HVGO then you add 1 or 2 or even 3 theoretical stages between the LVGO PA and HVGO PA stages.

 

To recover some gasoil that was left in the topped crude you could add a section above the LVGO PA consisting of a Top PA (2 theoretical stages) and a fractionation section of 2 or 3 theoretical stages between TPA and LVGO PA.

 

It is your design so you can play with the number of theoretical stages to achieve what you consider to be the optimum design.

[engeng]

Thank you very much PingPong, now it is very clear. My last question is: Should I use the efficiency of trays with actual number ?, I can calculate it from some literature, or there is a specific numbers for each sections of vacuum tower.?

[Technical Bard]

Kaes provides theoretical stages.  In Kister (among other references), you can find the O'Connell curve (1946) that relates fluid properties to tray efficiency.  Divide the number of stages by the tray efficiency.  In a Crude tower, efficiencies are on the lower end.

[PingPong]

Should I use the efficiency of trays with actual number ?, I can calculate it from some literature, or there is a specific numbers for each sections of vacuum tower.?

Once again: do not use trays in a crude vacuum column, except in the stripping section (below the feed). Use packing, and grid in the wash section (section above the feed).

Websites of Sulzer and KochGlitsch have brochures for such applications.