Tray Efficiency using Hysys 2

[imans]

Hi guys, I want to calculate tray efficiency using hysys from actual data. What I found was the tray efficiency was an input data not a calculated data. Can anyone give me suggestion to calculate the tray efficiency?

 

[jay165]

What I've done is a trial and error approach. Adjust the tray efficiency to match your actual input/output conditions. When you get a match, you've got your answer.

 

[Zoobie]

I have used jay165's approach and this works in a lot of cases (like when you want to try and match up HYSYS to your plant). Where this breaks down though is if you have widely varying efficiencies down the tower (unless by some miracle you have tray-to-tray input data). The differences can be great depending on what your tower is trying to do. For example, the experts tell me, that one of my towers has about 40% efficiency in the top and something like 60 to 70% in the bottom.

If all you are trying to do is to get your inlet, outlet, and energy streams to match than using the trial and error approach will work fine. The same holds true if you have a strong cause to believe that your efficiency doesn't vary a lot across the tower...or perhaps you can make an engineering guess. If the actual tray efficiency is what you want then other methods are probably better suited...I am interested in those as well.

As an afterthought, have you played around with the rating utilities in HYSYS? I can't recall if they do any efficiency calcs for you.

 

[sshep]

Imans,

Use this proven approach (assumed 1 feed, 2 product tower below) with any simulator to find overall column efficiency (theoretical stages / actual stages):

1) Specify a distillate and bottoms stream (flow and composition)
2) Mix the two streams in a mixer to use as tower feed
3) bring feed to plant conditions with a heater
4) send feed to a distillation column specifying plant values for tower pressure, reflux flow and temperature, and distillate flow (note that this is same flow as in #1 above)
5) Run sensitivity on number of theoretical stages- you can either keep feed location in proportion, or (if you believe section efficiencies can vary) you can run a 2 dimensional sensitivity on total stages and feed stage.
6) Pick the sensitivity result that gives the closest match to your distillate composition (bottoms will also be close due to material balance constraint). This is your number of theoretical stages.

Once you have used this approach, you will see how it works and can make various refinements to my method to suit your needs (including sensitivity on murphree tray or vaporization efficiencies if you are a fan of those).

The biggest stumbling blocks for new simulation users are: to try and use feed data as simulation input and expect product data to match plant (it may not be consistent)- steps#1 and #2 completely eliminates this problem; and to fail to recognize that we should use the actual plant values for reflux and distillate flow as simulation inputs- step#4 addresses this.

best wishes,
sshep