8 replies to this topic

[kwankwan]

Hi all.

I am having a problem with answering from a question. I am doing a short internship as a student in a petrochemical company, and I have a question about location of a condenser and a reflux drum in a distillation column.

I thought that a condenser are usually above a reflux drum as the flow goes naturally. But what I saw in the plant was opposite. It is a depropanizer and C3 component goes from the top of the tower to the condenser, which is on the ground. And the effluent from the condenser goes up to the reflux drum. But the drum is elevated by 5 meters above the top of the condenser. And I don't know why it is set up like this.

The condenser uses Cooling water as a refrigerant source from the underground. So I thought elevation between the condenser and the drum is the reason.

But do you think there is another reason for this type of condenser and reflux drum?

Thank you for reading this

[paulhorth]

Kwankwan,

It is true, I believe, that usually the condenser is above the reflux drum so that it can be free draining. However, the arrangement does depend on the control method used for controlling the column pressure. There are many varieties of control system for a couumn such as a depropaniser which has almost complete condensation of the overheads (with a small vapour stream of ethane etc). I have an article and a set of sketches for these control options but unfortunately I can't find this right now, i will get back to you.
The drum has to be elevated, to provide suction head to the reflux pump. If the condenser is lower than the drum, then the pressure in the drum must be lower than in the column, by sufficient to draw up the liquid. This pressure has to be controlled and this means that there is probably a vapour bypass line from the column to the drum with a control valve in it. The small flow through this valve of noncondensables plus a little propane will provide the lower prressure in the drum, the small amount of propane condensing in the drum at the liquid surface, which keeps the pessure lower.. I think that the condenser will run partly flooded, which controls the exposed cooling surface, and so the condensing rate and thus the column pressure.

If you can send a sketch showing all the connections between the condenser, drum, and column, and the control scheme, I will try to understand and explain it.

Paul

[fallah]

The condenser uses Cooling water as a refrigerant source from the underground. So I thought elevation between the condenser and the drum is the reason.

kwankwan,

I think, your above mentioned reason is the main one for installation of condenser on the ground.

Traditionally overhead condensers of Depropanizers/Debutanizers are Air Cooled type, which should be installed at elevated steel structure. In your case there might be some limitations to use air cooled condenser or cooling water has been used rather than cooling air.

Fallah

[pavanayi]

Kwankwan,

As Paul described it, this particular type of column pressure control is known as Hot Vapour Bypass. It gives a fast response to column pressure fluctuations, and the control is achieved by increading/decreasing the exposed condensing area by the liquid level inside the exchanger.
I can recommend you to read H. Kister's Distillation Operation, page 532 for a detailed description of this pressure control strategy.


Paul,

I may be able to help you with the attached document regarding different column pressure control strategies.

Attached Files

•  Distillation Column Pressure Control.pdf   2.06MB   514 downloads

Edited by pavanayi, 26 September 2011 - 08:15 AM.

[paulhorth]

Pavanayi,

Thank you for this download, it's just what I was looking for. I have something similar though much older.
I think the system Kwankwan has is No. 11 or perhaps No. 13. I sometimes struggle to understand that No. 11 will work, but I accept that it does. My problem is that if there is no flow through the bypass valve, then the pressures will equalise, but if there is flow, where does it go? That's why I think there has to be a small vent of vapour from the drum - but I have probably not understood something.

Paul

[pavanayi]

Paul,
Instead of me trying to articulate the idea and write a long paragraph, I am taking the help of Henry Kister.

Quoting Kister (Distillation Operation, page 532)
"The condensate must be subcooled so that liquid surface in the drum are cooler than in the condenser. This causes a difference in vapour pressure large enough to transport condensate from the condenser into the drum. When column pressure rises, the controller closes the valve. This reduces condensation at the drum surface and lowers the drum surface temperature, and, therefore, the drum vapour pressure. This enhances the vapour pressure diffference between the condenser and the drum, which inturn forces more liquid out of the condenser and into the drum. This exposes additional condensing surface in the condenser"

What you mentioned is the opposite of the above mentioned process.

I hope it helped.

[paulhorth]

Pavanayi,
Many thanks for the explanation by Dr. Kister - who I will not argue with. You are right, my attempred explanation was the wrong way round.
There's still something odd about what Kister says -

This reduces condensation at the drum surface and lowers the drum surface temperature, and, therefore, the drum vapour pressure.

.................I would have thought that reducing condensation would tend to keep the pressure up, and not to reduce it..... but there you are, I will accept it, this system clearly does work though needs to be carefully designed.
I have seen and used Fig 9 and Fig 10 with aircoolers, for this type of column, they are definitely easier to understand for me. These work by partial flooding of the tubes, but the aircooler is above the drum.

I have no experience to judge which of these options (Fig 9, Fig 10, Fig 11) is the best in operation.

Paul

[pavanayi]

Paul,
When the overhead vapours are condensing on the drum surface (which is cooler due to the sub-cooling within the condenser), the top layer of the liquid inside the drum is at a temperature between the subcooled temperature coming out of the condenser and the saturation temperature (when there is no subcooling) of the vapour.
The vapour pressure inside the drum will be lowest when no hot vapour condenses ( temperature of the liquid within the drum equal to the subcooled temperature), which equates to the control valve closed fully. At this state, the pressure differential is maximum between the condenser and the reflux drum and maximum number of tubes are exposed.
Similarly, when condensation needs to be reduced, the control valve opens to let warm vapors into the drum overhead, thereby increasing the pressure slightly. This leads to a reduced pressure differential and liquid backs up into the condenser, thereby reducing condensation.

[paulhorth]

Pavanayi,

Thank you for your clear explanation. I do understand it now.
I would express the transient change when the hot vapour bypass is closed, as follows.
The top layer of the liquid in the drum is warmer than the bulk liquid, but when the control valve closes, this warm liquid cools down by contact with the bulk liquid. The vapour in the drum is then in contact with cooler liquid, so a small amount of the trapped vapour will condense. Because the vapour valve is closed, this condensation reduces the pressure by a small amount, a fraction of a bar, which is enough to draw liquid up into the drum from the condenser. The drum level is held steady by the level controller acting on the reflux pump.

I find that it helps me to understand a control system if I can write down a narrative description of a transient change, then check if the opposite happens if the direction of the change is reversed.

When designing this system, it will be difficult for the process engineer to size the hot vapour bypass valve. He will need to estimate a flowrate, which has to be the rate at which the vapour will condense by contact in the drum during normal steady operation. I would find it difficult to make this estimate, I would guess it would be only a small fraction of the total vapour flowrate. The risk is making the bypass valve too large, which will make the operation unstable.

Paul