3 replies to this topic

[Mr.Frost]

Hi,

 

I am studying Chemical Engineering and I need to simulate the formation of propylene glycol with direct hydrolysis of propylene oxide with water. The reactions are:

 

PO + H20 --> MPG (main reaction)

MPG + PO --> DPG

DPG + PO --> TPG

 

with:

PO = Propylene oxide

MPG = Mono-propylene glycol

DPG = Di-proyplene glycol

TPG = Tri-proylene glycol

 

This work is based on information out of Ullmann's Encyclopedia of Industrial Chemistry. The process needs an excess of water and after the reaction the water needs to be stripped for the products to be seperated. The lean water is meant to reenter the reactor as a recycle stream. And there is my problem. Of course the recycle water is not 100 % clean. There are still some amounts of MPG, DPG and TPG which will react again in the reactor with the PO. Without the recycle stream the simulation works perfectly and I achieve the purity grades which are given. But if I try to recycle the lean water the simulation runs into nirvana. The weight stream of PO is given and fix.

 

Has anyone any idea how to solve the problem? I can variate the amount of water (recycle, makeup and purge stream if needed).

 

I really hope you can help me

 

And please excuse my English if it is not perfect. I am from Germany and still learning.

 

 

Viele Grüße,

 

Mr. Frost

[Bobby Strain]

Last time I looked at this system the water is recovered overhead from refluxed multi-effect evaporators. So it should be essentially free of glycols.

 

Bobby

[Mr.Frost]

Hi Bobby,

 

thank you for your answer. I am already working with dehydration columns to strip the water. It is for example possible to reach over 99 % water overhead. But the remaining amounts of MPG and DPG cause the reactor to accumulate the higher clycols and crash the simulation over time. There is a steady growing of the amount of TPG and a steady reducing of MPG and the makeup water stream needed.

 

How would you design such a refluxed multi-effect evaporater?

 

 

Edit:

 

The process looks like this apart from the fact that we only use one reactor and we do not need colums d and e.

(Source: Ullmann's Encyclopedia of Industrial Chemistry, Vol. 30, pp. 233-241)

 

 

Viele Grüße,

 

Mr. Frost

Edited by Mr.Frost, 04 June 2016 - 04:26 AM.

[Bobby Strain]

Just add a refluxed absorber on the water vapor from each evaporator.

 

Bobby