7 replies to this topic

[Ming Hooi]

I'm trying to find out the optimum diameter of a distillation column with parameters as below:

Type of feed, iso-propyl alcohol
Feed flowrate 4000kg/hr
Temp. 350K
Trays: Bubblecap
Tray spacing: 12 inches

Here's what I've done:

Using the formula

Vm = Kv (rhoL-rhoG / rhoG)^(1/2)
where Vm = maximum allowable superficial vapour velocity (based on cross-sectional area of empty tower), ft/s
Kv = an empirical constant, ft/s
rho are densities, with L and G denoting liquid and gaseous phase,

with tray spacing at 12 inches, I found that the Kv = 0.18ft/s

giving Vm=1.33m/s (by assuming the properties of ethanol - could not find the data for IPA)

the corresponding minimum diameter of column at 4000kg/h feed is 0.90m, and I added 10% safety margin to give 1.0m.

Why then some senior engineers are telling me that they've optimised the size at 1.27m?


References for calculations above:
1. Peters, M.S. and Timmerhaus, K.D., (1991), Plant Design and Economics for Chemical Engineers, 4th Edition, McGraw-Hill, Singapore, pp.656-657.
2. Perry, R.H. and Green, D.W., (1997), Perry’s Chemical Engineers’ Handbook, 7th Edition, McGraw-Hill, Singapore, p2-235.

Please advise.

[php78]

you need to take actual velocity which is approximately 0.7 times Vmax
and also need to consider downcomer area (approximately 10 to 15 %)

good luck
php

[siretb]

In order to properly size we need
the physical properties, that we can calculate from the temperature and the composition. We have that.
the vapor flow rate. This can be different from the feed rate (because of reflux). Please specify both the vapor and the liquid feed rate in the segment of column you want to size. We have not that.

Also 12 inches tray spacing is a very low figure. It may explain that extra safety factor has been applied. Anyway I cannot recommend such a low tray spacing. Since you write "optimum diameter" I guess you design a new column.
Two points: Why select bubble caps?
Why not play it safe, and go for 16" TS, at least ?

[weilongwu]

i think can use the packing column, siretb is right. you can use the simulation software(for example: aspen plus or pro/ii or Hysys) compute the actual vapor-liquid flowrate of column

[abhi_agrawa]

I will add to the above...
After having got the actual valor/liquid flow rate you can use some of the freely available softwares. eg. KG-Tower by Koch-Gilistch is freely available for download from their website. By this you should be able to use some of the commercial valve trays.
As for the tray spacing, usually 24 inches is used, for some applications, tray spacing of 18 inches can also be used. The tray spacing will be a strong function of the foaming tendencies of the system, so do not forget to use a proper foaming factor.
Hope this helps.

[Art Montemayor]

There are various impractical points discussed in this thread. I sense a lack of practical, hands-on experience on the part of some respondents regarding the process & mechanical design of distillation towers and want to only add some real-life facts that, while not very familiar to some, still are dominant enough to set the real criteria of tower design into an operable and maintainable light. I seriously doubt whether Max Peters, Klaus Timmerhous, and John Perry ever got the opportunity to try to install, inspect, clean, or even replace bubble caps within a 1 meter diameter tower with only 12 inches of spacing between trays. If they had, they would have re-written their famous discussions on the subject. The mere mention of such a diameter with only 12 inches of separation between trays is not practical. Today, even the choice of bubble caps is debatable if we are going to seriously contemplate the word "optimum". Their physical make-up and efficiencies are just some of the factors that have made bubble caps almost extinct in the process industry.

Unless you install more than one manway on each tray, there is no way on God's Green Earth that a human being is going to be able to tighten the cap nut on a bubble cap on a 1 meter diameter tower. It is not physically possible. Take it from a 68-yr old engineer who worked his way through Chem Engineering as a boilermaker. I've gone into some tight places and vessels, but that would be impossible. The basic equations used, together with the empirical constant employed in the Brown-Souders equation, have an accuracy as employed around 20-25 %. Instead of a "safety" margin, what should be applied is a contingency factor. A safety margin is used when you don't know what else to expect as a required factor affecting the sizing. A contingency is almost mandatory since the engineer knows in anticipation that the relationships applied have a noted inaccuracy in their results.

All the "design" textbooks like Peters & Timmerhaus are needed and required reading for understanding the basics of process design. However, these text books are guidelines - not bibles to apply without any engineering sense or decisions. Engineering has never been a profession where the simple application of some equations is sufficient to resolve a problem. Thank God! Common sense and good, experienced judgment is more important than merely applying equations. And this particular bubble cap tower shows us that value.

[abhi_agrawa]

Well I have commissioned a column with 900mm ID (35.4 inches). The tray spacing was 600 mm (23.6 inches) and had 30 trays. A tray spacing of 12 inches is definitely very low and it will not be possible to put the trays.

[niallmacdowell]

I think art is as usual, correct. If anyone is looking at how to design a distillation column, see Coulson & Richardson 6 for details, henry kister has two good books on the process as well, but there is no substitute for hands on knowledge