Distillation: top control temp changes with feed to hold spec?

[mjpetrag]

I have a distillation column with 3 packed bed sections, no trays. I did some historical analysis on the column and noticed that as feed to the column goes up (range of 3500-8000 GPH), the top control temp (underneath the top bed of packing) also needed to be higher in order to keep the top impurity level in range. The swing is from 112 C to 115 C at the high end.

Basically if the top control temp stayed at 112 and we go up in feed to say 7500 GPH, too much will flow out the bottoms and the overheads impurity level will be too low. However at 112 at 4000 GPH feed the right balance is made. If we were at 115 in the top at 7500 feed, the bottoms flow is minimized while top impurity is in spec.

The steam is controlled on a steam to feed ratio curve.

I had some theories as to why this was happening but wanted to hear any ideas as to why the top control temp cannot remain constant.


-Mike

 

[sshep]

Hey mjpetrag,

These kind of reproducible problems are often easy to troubleshoot with a hypothesis testing approach- i.e. brainstorming possible causes and testing to get data. Once you know the reason for the behavior, some control changes can result in practical solution. A sketch would be helpful.

Two possible reasons:
1) Tower pressure profile- when you load up the tower, the pressure drop across the top bed means the control point pressure is higher, so obviously the control temperature needs to be higher to get the distillate flow up.
2) The feed/steam "ratio curve" characterization is off- It is a common mistake to think you can just ratio feed/steam, but this assumes zero intercept which is not right. Steam is required even if there is almos no feed in order to load the tower and provide minimum traffic. The characterization is also sensitive to subcooled feed or other enthalpy assumption deviation.

Anyway, I think the forum will be able to help solve if you give a bit of a sketch, the system being seperated (concentrations), some detail about the feed condition (is it at bubble point, etc), instruments (panel and field points), and a few other details like that.

best wishes always,
sshep

 

[mjpetrag]

I think 1 is the likely scenario. the temperature in question is underneath the top bed of packing.

Originally when I looked at it, I thought since it was packing that the pressure drop across 1 bed of packing would not significantly alter the temperature. Looking at the VLE data for the material in question, the difference in Tsat from a 15 mmHg change is pretty significant

Psat at 40 mmHg is 100 C
Psat at 65 mmHg is 110 C
Psat at 100 mmHg is 120 C

I'll have to look at this again to make sure the numbers make sense tomorrow.

-Mike

 

[mjpetrag]

*Tsat not Psat

-Mike