RO/NF combination for fufural concentrating

[sero2000]

Hi there,

My group is working on a final semester project that involves the design of a small fufural (C5H4O2) plant. Pretty much most commercial processes uses distillation after the reaction process and then it is further distilled to purify the fufural which results in a costly process

Feed-> Reactor -> Distillation -> Purify (Distillation, Liquid-liquid extraction ect)

We are looking into removal of the water after the reaction process to first concentrate the stream before sending it into purification stage where we will be using a distillation column. We are thinking of using the concentrates from the NF and RO processes and feeding it into the distillation column since the main thing that is affecting the cost is the large amount of steam needed to vaporise the water-fufural mixture that is present after the reaction process

Feed -> Reactor -> NF + RO -> Distillation

Does anyone who has any experience in membrane separation or fufural could give any advice on whether this is even feasible?

Im not sure if i posted in the correct section so i do apologize if i did so wrongly.

 

[Bystander]

https://en.wikipedia.org/wiki/Furfural
If the solubility stated in Wiki is correct, enthalpy required for water vaporization isn't going to be your big heat requirement.

 

[sero2000]

https://en.wikipedia.org/wiki/Furfural
If the solubility stated in Wiki is correct, enthalpy required for water vaporization isn't going to be your big heat requirement.

hmm so you're suggesting we focus on the reaction process instead?

 

[Bystander]

reaction process instead?

Not necessarily --- re-reading, you're proposing RO/NF as an alternative to "brute force" distillation. Yes, good idea. Solubility in water is ~ 1 molal, should keep pressures needed for RO manageable, IF the reaction/feed/by-products don't interfere too much or contribute too much other water soluble material.

Your project --- probably enough from the "peanut gallery." Update us if time allows.

 

[sero2000]

Hmm i looked at a study that showed that concentration of furfural was done using membranes with rejection of furfural of over 92%. However, it also showed that acetic acid (by product) also was rejected at around 70%. And in another study when they increased the trans-membrane pressure the acetic acid retention was 15%

MW of acetic acid is around 60 while for furfural it's around 96. General RO porous membranes (from GE) that I looked at have around 100 Molecular weight cut off. In this case, do you think it is better that I use a non-porous membrane and look into solution diffusion models since the feed that is coming out of the reactor is expected to be 90% water, 6% furfural and 4% acids

 

[Bystander]

solution diffusion models

Might be interesting, but probably not all that helpful as far as cutting distillation costs.

90% water, 6% furfural and 4% acids

... plus just enough solids to be a real fouling problem. Cross-flow vs. diffusion --- ?

Do you get to do any actual measurements/experiments, or is this "commit to design" and hope for the best?