9 replies to this topic

[Devoiex]

Hi, I'm a university student doing my final year project in sizing a De-methanizer, De-ethanizer as well as a De-Propanizer.

I've been told to do a rough sizing of the distillation columns using heuristics. I do know that a number of heuristics that could help me with my sizing of my columns such as optimum reflux ratio should be 1.2 times of the minimum reflux ratio and etc.

But one thing that has been bugging me is that if there is a rule of thumb whereby i could get a rough estimate of my column diameter using heuristics rather than going through calculating the flooding vapor velocities and the Liquid-vapor flow factor by using

Uf = k1*sqrt((rhoL - rhoV)/rhoV)
(Souders-Brown's equation)

and



I have been reading through a lot and none of them have got any shortcut designs for distillation columns. Is there any procedure existing in the industry for distillation column in terms of shortcut design? Or i am unable to do any estimates with regards to the column diameter but to only resort to the above equation?

On another hand, I would like to ask anyone who have gone through any actual fractionation designing that is it the practice to specify the hole size for trays within the column as well as the weir height to acquire trays as designed or is there a standard set of weir height along with hole size that the manufacturers have that it would be cheaper to purchase instead of custom making one. I've been going through manufacturers like koch-glitsch for their product catalog but to no avail of any given specifications.

Oh yes, i'm dealing with a multicomponent fractionation.

Hope that someone can give me some insight with regards to the above.

Cheers.

Edited by Devoiex, 25 April 2010 - 03:51 AM.

[Zauberberg]

In particular: NO, there are not shortcuts in distillation column sizing because the dimensions will strongely depend on what type of internals (trays/packing) and flow pattern (single or multipass, dual flow etc.) you will choose for your tower.

In general, there are several shortcut methods and you can look into the attached document for more information (page 7): http://www.aiche.org.../heuristics.pdf

Also, for evaluating different designs quickly, you can go to: http://www.koch-glit...re/default.aspx and download KG-Tower software. By changing the input information (tower internals and flow regime) you can get a realistic picture of the final design.

[wojtar]

Well, as far as I know Souders Brown is widely used for estimation of column diameter. You can download KG-Tower software from Koch-Glitsh website or Sultzer's Sulpack. I don't remember but i think they have sieve trays calculations. I supose there are standard layouts for sieve trays and different column diameters but the hydraulics must be checked. Weir height is very important for column hydraulics and tray pressure drop. I supose taht for sieve trays it will be just few milimeters, say 15 mm - it must be checked. By the way - that Souders Brown graphs can be turned into equations which can be solved using numerical methods. You could search deeper the internet or read simulation software helps for that.
If your column is intended to work in reality and the end-client is known, Koch-Glitch will prepare hydraulic calculations for you, but if it's university work - I don't think so.

Regards.

[katmar]

I would regard the Souders-Brown correlation as already being the short-cut method. If you have experience in a particular application, and you use standard tray spacings, you could probably use the vapor velocity to get a guesstimate for a new column. As wojtar has pointed out there are versions of the S-B equation that have been expressed numerically. See Fair's method in Perry 7th Ed.

If you are going to use a major supplier like Koch-Glitsch then they would be more interested in the actual components and pressures than in your (or my) opinion of the required hole size and layout. They will make their own selection based on fouling tendency, corrosivity etc. If you are going to have the trays made up by a general fabricator then you would just contact them and ask if they have standard dies for punching the holes. These will tend to be 1/4", 3/8" etc or integral millimetres depending on whether you are in the USA or not. In my experience there is not much saving to be had in the DIY method.

The weir height will depend mostly on the pressure drop you have available. Again, the fabricator would advise but as a starting point you could guess in the 25 - 50 mm range. I have not seen weirs as low as wojtar's 15 mm, but perhaps that is possible in vacuum applications. The lower the weir the more sensitive the tray is to out-of-level tolerance.

[Devoiex]

Thank you so much for the above replies. I appreciate them.

On another hand, i would like to ask is that in order to obtain the number of theoretical stages for a multicomponent fractionation column i have read that it is possible to use McCabe-Thiele method. I have been using it for binary mixtures since a long time but as for multicomponent fractionation, i have got no idea how to go about doing it.

Is there any other method in obtaining the column stage estimation? I was thinking of approaching the problem by using trial and error method in hysys by slowly decreasing the number of stages and noting down the composition of my required stream but i have been reading too much of people saying hysys is unreliable and i do not know if this method would actually work out fine.

I do know that it is possible to obtain the minimum amount of stages required for a given fractionation problem by using the Fenske Equation below.



I would like to know that is there any other way to obtain the theoretical stages required for a multicomponent fractionation column to obtain a flexibility of calculation as stated above, the Fenske Equation is applicable for total reflux thus i assume it is not for partial or full reflux

Any help would be greatly appreciated. Cheers.

Edited by Devoiex, 25 April 2010 - 02:30 PM.

[Zauberberg]

Assuming we are speaking about conventional distillation tower (overhead and bottom products only), you could use a simplified approach where light and heavy components are the same as in the case of binary splitter. I have never used it though, so I cannot tell you how accurate it is.

In any case, whatever design case you have in your hands, it has to be confirmed by the vendor. It's the normal, common way of doing distillation design: you prepare the set of input data, product specifications, and available heat sources and heat sinks (= column pressure), and vendor will come up with an optimized design. I would definitely follow that path if I were at your place. Sulzer and Koch-Glitsch are usually quite keen to address almost every technical query and feed you with preliminary design information.

[Devoiex]

Thanks and sorry for so many questions that I'm posing over here but does anyone have got any idea how to find the K-values or the vapor-liquid ratio within the FENSKE equation?


where

α = the relative volatility of the more volatile component i to the less volatile component j
yi = the vapor-liquid equilibrium concentration of component i in the vapor phase
xi = the vapor-liquid equilibrium concentration of component i in the liquid phase
yj = the vapor-liquid equilibrium concentration of component j in the vapor phase
xj = the vapor-liquid equilibrium concentration of component j in the liquid phase
(y / x) = K commonly called the K value or vapor-liquid distribution ratio of a component

I have got a working copy of my hysys simulation for 3 fractionation column but i have no idea where or how to obtain the relative volatility from hysys. I can only find the K values within the pipes but not within the fractionation column. Cheers.

[Zauberberg]

I don't have HYSYS with me at the moment but K-values are definitely there in the main column window.

[Zauberberg]

Here it is: Column Window --> Performance --> Plots --> K-values (table). Attached.

Attached Files

•  K.jpg   135.6KB   65 downloads

Edited by Zauberberg, 26 April 2010 - 09:42 AM.

[Devoiex]

Here it is: Column Window --> Performance --> Plots --> K-values (table). Attached.

Cheers! thanks a lot! you're a savior and pardon me for my incapabilities in hysys! I'm done with all the preliminary stage of fractionation sizing thanks to all the above useful comments. Just in case anyone else who is searching this forum bumps by this thread of mine I'd like to say that a few books listed below are really good references in sizing multicomponent fractionation.

1) Applied Process Design For Chemical and Petrol chemical plants vol.2 - Ernest E. Ludwig
2) Coulson & Richardson's Chemical Engineering vol.6 - R. K. Sinnott
3) Distillation Operations - Henry Z. Kister

Useful Heuristics
http://www.aiche.org.../heuristics.pdf
http://www.chem.mtu....pmentSizing.pdf

General steps in sizing a distillation column
http://www.sammichae...ftp/class14.pdf