14 replies to this topic

[laobill]

Hi, I'm designing a demethanizer column as part of my final year project, could anyone tell me please, if I can McCabe-Thiele method to calculate the number of trays required and if I have to design it as if for binary component as there are just 2 exit streams, even though there are multicomponent (C1 to C10). Also, how would I design the top part of it please?

[abhi_agrawa]

Laobill,

A Demethanizer, as the name suggests is used to remove Methane from the stream. You can use the McCabbe-Thiele Method to get an initial estimate of the number of trays you would need. However, for the complete design you would probably need to simulate the same on some simulator.

Have you decided on the pressure of the column? This is very important, because the normal boiling point of Methane is about 161.1 deg C and of Ethylene (the heavy key) is -103.8. Sice these temperatures are quite low, you have the following options:
1) Increase the pressure to increase the condensing temperature, or
2) Get a methane refrigeration system.


Hope this helps,
abhishek

[laobill]

Hi abhishek,

Thanks for your help, there are refrigerator and turboexpander before the demethanizer and I have fixed the outlet stream pressure from the turbine to 13 bar, which would be the same pressure the column would be working at. At that pressure, the 2-phase mixture (C1, C2 and C3) temperature going into the demethanizer is -72 deg C, with the liquid phase (mostly condensed C2 and C3) acting as reflux. There are total of 10 components going into the column (C1-C10). Would I have to calculate the column pressure based on methane dew point or I can fix it at 13 bar?

regard

laobill

[abhi_agrawa]

Laobill,
To start with you can fix the column pressure at 13 bar.
From your description I am unable to understand the process flow. It would be good if you can provide us with a schematic/PFD. The demethanizers that I have seen usually have multiple feeds. Since the demethanizer overhead is quite cold, it is used to chill the demethanizer feed. All the chilling is not done in one shot, but rather in several exchangers, and between each exchanger there is a separator drum, from which the condensed liquid is taken out and sent to the demethanizer. This helps in reducing the duty requirement for the demethanizer. Also in this way you can separated some heavy stuff (C6 - C10), which need not be sent to the demethanizer. However all this depends on your feed composition.

Abhishek

[laobill]

Hi Abhishek,

Thanks for your help. I have applied a simple version of the PFD with material balance. I just need help determining the number of stages, the column diameter ( top bell shape and bottom half) and the feed location.

Regards

Laobill

[abhi_agrawa]

Laobill,

For the PFD you have shown, I am not sure if McCabbe-Thiele diagram can be used for the determination of the number of feed stages. This is because the reflux and the vapor product from the column are not in equilibrium. If you have an access to any simulation software, then you can probably employ the brute force method. You can assume some number of stages to start with, optimize the feed location, then increase or decrease the number of stages and optimize again an so on. Then tabulate the results and choose what seems to you to be the best.

In the sketch that you have attached you show the feed location and the swage at different locations. Please note that this will not be the case in real world. The swage (ie. transition of diameter) will be at the feed location. The column diameter is fixed by the vapor-liquid traffic in the column and this will change most significantly at the feed location or at a side draw etc. You have shown the column to be a trayed one. So you will need to generate the vapor-liquid traffic in the column, then you will have to do the column sizing. The column sizing can be done by using any tray vendor software (those by Koch-Glitsch and Sulzer are available on their websites for a free download). However I am not sure if a packing would be better. You may send you loads to the vendor and they will also give you the diameters.

Hope this helps.
Abhishek

[laobill]

Thanks Abhishek, that helps a lot. However, would you say the top part of a Demethanizer column is a rectifying section or rather a gas desorption section/column? I have used Pro/II simulation software many times to solve this but it wouldn't work probably because like you said, the reflux and vapour product are not in equilibrium. Would you normally solve for the top part of the column as a different column or just as a rectifying section of a distillation column?

Thanks,
Laobill

[abhi_agrawa]

Laobill,

I do not see why the column would not solve. My tips for the same:

1) Give some initial estimate for the reflux to the column. This will help the column in reaching at some solution in the first pass, this increases the probability of reaching a solution.

2) Ensure that you are not using too tight tolerances. If you wish to use a tight tolerance, get the column to converge first at loose ones and then give initial estimates for the temperature profile for the column. You may have to give composition estimates also for solution at tight tolerances.

3) Make sure that in the column setup, you simulate it as a column with no condenser.

Despite the configuration, I'd still call the section above the feed as the rectifying section. I'd also say that you should check the column pressure. With the kind of flow scheme you have, I feel that it may be better to go for a higher pressure. Just check if that is feasible, because the column top temperature would be quite low (of the order of -120 deg C).

Hope this helps,
abhishek

[laobill]

Hi Abhishek,

The simulation works now after I specified an initial reflux and exit streams compositions. Thanks a lot, I really appreciate your help.

Laobill

[smalawi]

Hi,

i think chemsep is offering a free edu demo that you can use. it have the ditillation diagram and othe rthings that help visualise the problem,

cheers,

sm

[laobill]

Hi abhishek,

Just wondering if demethanizer design procedure would be the same as in normal distillation column! I'm still not totally sure if the only two differences between them is the absence of condenser and the fact that the top section diameter is larger than the bottom section's? Could you tell me what other factors I have to consider when specifying the column hydraulics please?

Regards,

Laobill

[abhi_agrawa]

Laobill,
Well the design procedure for Demethanizer would be same as other columns. Things to be taken care of:
- This is expected to be mildly foaming service, if you have trays then you may have to de-rate the jet flood and downcomer flood
- Usually packings are preferred in this service.

The diameter of the tower would depend on the vapor-liquid traffic. Usually in towers like demethanizer, the vapor traffic controls the diameter. Thus if the vapor load in sufficiently higher in the rectifying section as compared to the stripping section then the diameter in the rectifying section will be more than the diameter of the stripping section. However, if the difference in the diameters is not much, then you can consider having a column of single diameter. Whether to have two different diameters or to keep the same is usually governed by economics.

Hope this helps,
abhishek

[want2know]

Hi everybody,
I need ur help...
Now i'm working on my design project to produce 6000 ton/day of sales gas. My problem is to determine the appropiate temperature and pressure for the feed and the product streams of my demethanizer, deethanizer, and depropanizer. What actually the criteria involved to set all these parameters?

[laobill]

Hi,

As you can see from the posts above, the feed temperature into the demethanizer will have to be about -120 degree Celsius and the pressure about 20 bar to extract about 100% of the methane. You can use quest consultant software (www.questconsult.com) to varies the conditions of the streams and see the compositions that best suit your outlet conditions.

hope this helps

laobill

[want2know]

[Hi laobil...

Thanx for da help....
i got a lot of informations from that website....