Question on refinery distillation column behaviour

I have a debutanizer column with feed consist of C1 hydrocarbon up to

180degC FBP Heavy Naphtha cut. The column is meant to split light hydrocarbon which ends up at the top and naphtha (light and heavy naphtha) down at the bottom. Light hydrocarbon at the top is basically C1 up to LPG components. LPG is drawn as reflux drum distillate while C1 and C2 is released as offgas from the reflux drum.

The column overhead cooling is provided by 2 fin fan cooler and a water cooled trim cooler before condensed vapor enters into the reflux drum. Some of the liquid from the reflux drum is routed back to the column top tray as reflux beside being drawn as the distillate.

Control of the top LPG quality is by way of controlling tray 11 temperature.

Theoretically when reflux temperature is low, the tray 11 temperature would drop. However, an instance happened when a sharp drop in fin fan and the water cooled exchanger outlet temperature raised the tray 11 temperature drastically. At the same time, the reflux drum level also drops.

Can someone explain how this could happened?

Reply to
The Romanov
Loading thread data ...

The Romanov a =E9crit :


Is your column a real one or design case ? Are you making a dynamical simulation or a static one?

you can not directly play on the reflux tempreature, it is fixed by the pressure and the composition of the distillate. reflux temprature is not a degree of freedom. However to discuss about this case, it is important to answer my first question.

Reply to

TheRomanov a =E9crit :

In fact starting the new fin fan, created a perturbation to the system. so you are in dynamic regime. I'm have not really high expertise in distillation dynamics, but it is well known that the process can show wath we call an ''inverse response'', this is due to the non linearity of the system and the coupling of the control loops. But normaly after this transitional period the temperature profile should go on the right way. so the question did you notice just a transitional temperature behavior or it stay a this new level after a period of time (steady state regime).

A high temperature of the tray means that the composition of the tray changed at constant presure. More heavy component are on the tray 11.

another point is the pressure. If the pressure raised this will also increase the temperature profile at constant composition. did you check if the pressure was constant.

However we can make a simulation and this will help to understand the problem.

Reply to

Can you describe the control system for the column overhead? The actions of the controllers may have contributed to the phenomenon.

I presume the column pressure is controlled with a valve on the light gas stream taken from the reflux drum. (In my mind I'm picturing a column with no bypass around the condenser, by the way, since that wouldn't make sense here. But please confirm that.) I'm guessing the column pressure dropped due to the increased cooling, and the offgas rate was pinched back by the pressure controller. How fast did the pressure controller respond? If too slowly and the column pressure dropped sharply, there could have been an increase in the vaporization in all column sections. That vaporization could have brought heavy components up the tower, increasing the tray 11 temperature. Looking at responses of other temperatures in the column (such as the total overhead vapor) may help confirm or refute this.

How are the drum liquid level and tray temperature controlled? One probably adjusts the reflux flow, and the other the LPG draw off. But which way were they lined up when this happened, and how fast did they react? My guess is that the effect of decreased reflux temperature (and increased flow, whichever control arrangement you had) on the tray 11 temperature would be quite slow relative to a sudden increase in vaporization carrying up heavy components.

This might be all wrong, though, if the pressure control was tight and the liquid flow changes were rapid. In that case post more info about the dynamic responses of the flows and pressures, and maybe someone will figure it out.

The Romanov wrote:

Reply to
Rob Brendel

My guess is that two things happened, perhaps by coincidence, at the same time.

When the reflux is cooled below its condensation temperature (at the current pressure), there will be further condensation of vapor at the point where the reflux enters the column. This condensation increases the amount of reflux in the column. The vapor condensation due to cool reflux is known as "internal reflux".

An increase in reflux (due to internal reflux) should decrease the temperature of the upper trays.

Perhaps something happed at the same time, such as an increase of heavier material in the feed that would increase the temperature. The decrease from the cooler reflux must have been less than the increase resulting from some other effect.

Sometimes (not often, but sometimes) when the operator or control system makes a change that should have one effect some unrelated thing happens that has the opposite effect.

Once, when operating a column I increased the reflux flow. What I didn't know (at the time) was that an upstream change increased the heavy stuff in the feed, causing the purity of the top (by analyzer) to fall in spite of my reflux increase. I was confused until I found out (from analyzer data upstream) of the coincidence.


The Romanov wrote:

Reply to
John Shaw

Have you considered some of these other possibilities:

  • Instrument drift or failure
  • Upset in feed stream concentration
  • Other process interactions

You would expect adjacent tray temperatures to move in the same direction as Tray 11. If tray 11 is not moving similar to, for example, trays 10 & 12, then this could point to instrument failure.

An upset in feed concentration could also affect the Tray 11 temperature. Does the column data show any evidence of this?

Process Interaction Mapping would also identify any strong correlations to your Tray 11 temperature loop. See this link for more info:

formatting link

The Romanov wrote:

Reply to

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.