Distillation process control

[Fahad3608]

Hi everyone,

I have a question regarding the distillation process control, we all know that distillation process is used to separate two key components. The temperature and pressure gradient are high at the bottom and low at the top, (bottom to up). However, my question is, why there is a specific tray that need to be controlled in terms of temperature regardless the other trays ? (i.e. tray No.7 temperature need to be maintained between 45 C and 48).

Appreciate your contribution in advance.

Thanks !

 

[don1980]

First, recognize that the temperature profile is a direct reflection of the composition changes. When the pressure is held constant, a slight change in the temperature means there's been a slight change in composition. If you plot the T profile in the rectifying section and the stripping section you'll see that neither one is linear - the concentration of the components is not changing linearly. In one particular set of trays you'll see a relatively high temperature change from one tray to the next (a steep slope in the T curve), while in other trays the T change is relatively flat. Those areas where the temperature (concentration) is changing steeply are the high-sensitivity zones, and those are the zones where you generally want to do your temperature control. In the process design stage, the simulation reveals these areas of the column, and these are a the areas of the column where the temperature probes are concentrated.

 

[jari001]

don1980 is right on. I'll just add that temperature control may be important at the feed location, as an example of one try needing temperature control, if the feed is conditioned properly (i.e. heating or cooling to desired temperature). You would want to monitor the areas where there is the greatest delta T from tray to tray and the feed location because the separation may not tolerate large fluctuations in temperature or you might start reactions at high temps if not monitored(oh boy!).

 

[sshep]

Hello Fahad3608,

If you have access to process simulation tools, you should check out the temperature and composition profiles for one of your distillation examples. Fix the reflux and distillate at plant values, and look at the profiles and splits for different tower heights. A short column will have continuous changes at each stage. A tall column will have invariant sections of temperature and composition- i.e. at some point taller doesn't help. Also observe that distributed components (volatility between keys and appearing in both top and bottom) will have peak concentrations at some stage in the middle of the tower. Observe how the profile and product composition changes when reflux or distillate are changed. This profiling exercise will help give insight into steady state distillation behavior.

A typical technique for finding a good inferential temperature control point is to make a perturbation of the material balance (start with +/-2% distillate draw) while keeping the reflux fixed. The composition of the products will change (how much depends on how tight the split), and so will the column temperature and composition profiles. Plot the stage vs temperature profile in Excel for the three material balance cases (base, plus, and minus). The stage in the column which shows the greatest sensitivity to temperature is a good candidate for the inferential control point.

At one time I was planning to create a set of FAQ's for distillation modeling, but back then the ability of engtips to embed links to pictures was unreliable. My first FAQ attempt (determining # of theoretical stages using plant data) did not get a good reception, however, attached is a link to three slides that illustrate the concepts above.

Good luck,
Sshep

 

[plantprowler]

@sshep

Could I request you to post the FAQ or any other info you could about distillation modelling. It is a sorely lacking area of advice in general.

There's very little good quality material online. I'm sure many others will greatly value your tips even if in the form of a simple post.