Troubleshooting of WFE - (Wiped Film Evaporator)

[Rachit]

TL;DR

Continuous boiloff operation in WFE.

Feed Details :
Component X : 80% ( Boiling point : 105 deg.C)
Component Y : 20% ( Boiling point : 256 deg.C)
Feed Rate : 600 kg/hr
Heating medium : Steam @ 145 deg.C
operating condition : 50 Torr, 90 deg.C ( Based on Aspen Flash Simulation).

Desired Outcome :
TO remove 99% of component X from Component Y.

History :
Earlier this operation was performed in batch mode. Where all of component X was boiled off.
During this, Bottom mass temperature was observed to go upto 120 deg.C uder 50 torr of vacuum.
No component Y was observed in the top distillate - throughout the operation.

Current Scenario :
Observations in WFE :

1. In top distillate component Y is found ~10%.
2. In bottom residue component Y is found ~80%.
3. At 50 torr vacuum, operating temperature is only 40 deg.C.
4. Upon increasing flowrate, component Y is found less in top distillate.

What should be done to minimize loss of component Y in top distillate ?

 

[Chestermiller]

Have you tried to model this operation? Is the solution ideal so that Raoult's Law can be applied.

 

[Rachit]

Have you tried to model this operation? Is the solution ideal so that Raoult's Law can be applied.

The solution is not ideal, but I have modeled the operation in Aspen plus and component Y in top distillate is ~0.5% w/w.

 

[Chestermiller]

The first step is to determine why your model is not matching the observed split. What are your thoughts on this? How do other observational measurements compare with model predictions? How confident are you in the VLE part of the model?

 

[horacio torres]

My friend

Aspen Plus has a set of properties packages that will approach the simulation to the real process. It has a tutorial to select the best ones. However, you have to select PR if your components do not have a polar behavior.

To optimize the process, it could be done with the process data. Of course, anyone needs to know X and Y, to have a view of the distillation equilibrium of the mix. Simulation is a tool, just that; but trained process engineers should do it by analyzing the field process data.

Horacio