use of recycle to improve reactor performance

[mikefot]

Dear All,

I am interested to ask a general question about the use of recycle to improve the performance of a chemical reactor.

If you operate a process in a single pass mode ie without using recycle and obtain a particular reaction conversion and spacetime yield (with a particular residence time) then
it might be useful to try recycle to see what the effect would be on the reactor performance.

I had originally thought that the best reactors would operate with good conversions and spacetime yields without needing to use recycle.

I tried asking one of the artificial intelligence LLMs about this.

It suggested that in some situations you could use recycle and get lucky with it and see both a substantial increase in conversion percentages and STYs and that it
was good engineering practice to take advantage of the recycle permanently in operating the reactor if it turned out to improve the reactoe performance.

For me a problem with the AIs like chatgpt and bard etc is that they can make mistakes and get things wrong etc.

As a result I thought to post this question here to check out people's opinion on this on the site here.

I hope the answer is of general interest enough to be useful to others not just me.

Regards

MF

 

[georgeverghese]

Reactors with high residence time are expensive in many cases. Many industrial reactions are reversible, so you never get full conversion anyway. Increasing temperature to speed up kinetics is an alternate to high residence time, but then side reactions and thermal decomposition of the reactants may often occur. In exothermic reactions, you often get better conversion at higher temp due to faster kinetics at a given pressure, but at the expense of lower thermodynamic equilibrium constant. Good catalysts and catalyst promoters help to increase conversion in reversible reactions at lower temperature and/or pressure, at lower residence time and with high selectivity for the desired product.

 

[mikefot]

Many thanks for replying to my post. Your comments and information I am finding in online searches and use of the AI LLMs is beginning to converge and gel together now.

In theory if you were operating a process with a short residence time which was irreversible you could benefit from using recycle in such a case especially if you could run it in a way that
would consume most of the reactants ie not leave too much unreacted material in the outlet stream.

It ia helpful to think about these things.

Cheers

MF

 

[Rocco8]

Set your recycle loop for different flowrates and see how the conversion improves at each setting.