Jump to content



Featured Articles

Check out the latest featured articles.

File Library

Check out the latest downloads available in the File Library.

New Article

Product Viscosity vs. Shear

Featured File

Vertical Tank Selection

New Blog Entry

Low Flow in Pipes- posted in Ankur's blog

Modelling Emcd Trays In Dynsim/unisim

distillation column distillation tray tower refining sepration emcd sieve tray weir weir length

This topic has been archived. This means that you cannot reply to this topic.
7 replies to this topic
Share this topic:
| More

#1 joyd

joyd

    Brand New Member

  • Members
  • 4 posts

Posted 10 February 2014 - 12:32 PM

We are trying to model EMCD trays for  a relief loading study using dynamic simulation . We have used baffle trays to model them with a very high weir length fraction (12.17) . That is the number we got  from the data sheet , diving the total weir length by the tray diameter.

 

 I wanted to know if anyone had any experience in modelling a EMCD tray with the conventional options available such as sieve, valve , baffle etc. 

 

 One advantage that the baffle tray offers is the low pressure drop , since this is a propylene splitter and we have a large number of trays .

 

 

Kindly let me know if more details are reqd. Such as the data sheet etc .

 

With Regard...



#2 PingPong

PingPong

    Gold Member

  • Members
  • 1,312 posts

Posted 10 February 2014 - 03:41 PM

ECMD (not EMCD) trays are definitely not baffle trays. They are produced by UOP. Google it and look at the pictures.

 

Baffle trays are just stupid plates, without sieve holes or valves, whereby the liquid simply runs over the edge and the vapor zigzags through the liquid curtains. That's why the pressure drop is very low, but so is the efficiency.

 

I suppose you can consider ECMD trays as sieve trays with many downcomers. You should find out the design details of the trays that are used in that propylene splitter such as number and width of the downcomers, outlet weir height, ........



#3 Bobby Strain

Bobby Strain

    Gold Member

  • Members
  • 2,806 posts

Posted 10 February 2014 - 05:37 PM

Did you talk with UOP? And, probably more important than the tray type is the simulator model. Relief flow modeling is a very demanding task.

 

Bobby



#4 joyd

joyd

    Brand New Member

  • Members
  • 4 posts

Posted 11 February 2014 - 01:55 AM

Hello PingPong /Bobby

 

   We are getting the reqd. separation with the baffle trays as well . In the model we have considered the efficiency as 1 . And tuned the conductance across the tower to get the rqd. pressure drop .

 

 

  The only thing I am concerned as this point is how will the flow dynamics be affected when the scenarios happen  ie. cooling water failure and the power failure

 

In this case the levels fill up very fast . and the column runs in the flooded condition for some time before relieving. I have the design details ,such as weir height , weir length fraction . But do not know how to read for the dowcomer width .

 

Will anyone be able to guide me in case I upload the datasheets ?

 

Thanks

 

PS: I have contacted UOP , haven't heard rom them yet



#5 PingPong

PingPong

    Gold Member

  • Members
  • 1,312 posts

Posted 11 February 2014 - 06:11 AM

We are getting the reqd. separation with the baffle trays as well . In the model we have considered the efficiency as 1 .

If you simply specify 100 % efficiency for the trays then, for the separation, it does not matter what type of trays you use. In reality no tray has an efficiency of 100 %, not ECMD trays (maybe 70 %) and certainly not baffle trays (maybe 20 %).

 

I don't really understand what it is that you are trying to simulate, but it seems to me that in a dynamic simulation the liquid holdup on the trays matters. This liquid holdup depends on the tray design: there is an amount of liquid on the tray itself, depending on outlet weir height and gas fraction (bubbles) in that liquid; and there is liquid in the downcomers. The liquid composition on each tray will moreover be different.

 

The only thing I am concerned as this point is how will the flow dynamics be affected when the scenarios happen  ie. cooling water failure and the power failure

 

In this case the levels fill up very fast . and the column runs in the flooded condition for some time before relieving.

What levels are you talking about? The one in the bottom? Do you expect that the column will fill with liquid before relief starts? I don't see why.

If cooling water fails (and consequently condensor fails) while heat input in the reboiler continues, the column pressure will start rising and the relief valve opens as soon as its setpoint is reached.


Edited by PingPong, 11 February 2014 - 06:13 AM.


#6 joyd

joyd

    Brand New Member

  • Members
  • 4 posts

Posted 11 February 2014 - 12:20 PM

 The Propylene Splitter consists of two towers . One is the rectifying section and the other the stripping section . The sump of the stripping section has the re boilers .

 

The inventory on the tray is determined by the weir height . Which is the same as the data sheet. And we are assuming a clear liquid height. And again we are getting the pressure drop by tuning a conductance factor. 

 

What I am seeing is with cooling water failing .The pressure rising , the sump levels rising in both towers. For the stripper section , the steam flow to the reboiler is cascaded to the level controller , so there is more evaporation.

 

The level of the rectifying columns is maintained by transferring more liquid to the stripping column . This happens via a pump and a level controller valve in the discharge

 

The vapor going to the overhead condenser cannot be cooled so the pressure rises , you are correct . There is a reflux pump that transfers the reflux from the condensate drum to the tower top tray .

 

The vapor in the rectifying  tower condenses due to high pressure that is my understanding .  Then the level in the sump starts increasing . 

 

So in your experience what is the expected behavior for such systems ?

 

Regards and greatly appreciate your inputs 

 

Joy



#7 Bobby Strain

Bobby Strain

    Gold Member

  • Members
  • 2,806 posts

Posted 11 February 2014 - 01:48 PM

It would be enlightening if you presented the results of your modelling. And also the estimated relieving rate using more traditional methods that don't rely on dynamic simulation models. It might also help to know your objective for the dynamic modeling effort. I presume this is an existing installation and not just an academic exercise.

 

Bobby



#8 PingPong

PingPong

    Gold Member

  • Members
  • 1,312 posts

Posted 13 February 2014 - 01:51 PM

The vapor in the rectifying  tower condenses due to high pressure that is my understanding .  Then the level in the sump starts increasing .
There is only condensing if the condensing heat is removed, and that is not the case when cooling water fails.

 

I don't know what the operating conditions of your propylene splitter are, but normally such column has a high reflux ratio, so the reboiler duty is normally much higher than the duty required to completely vaporise the feed to the splitter. Therefor I don't see how the level(s) in the splitter bottom(s) can increase if the reboiler stays in operation.

 

Without knowing flowrates, compositions, duties, number of theoretical stages, operating pressure, relief valve setpressure, reboiler medium temperature, et cetera, it is for outsiders difficult to picture the problem.






Similar Topics