Reactor Design -- Plug Flow Reactors

[JeweliaHeart]

As part of a research project, I've been asked to design an industrial scale model for the synthesis of titania nanowires. These are currently being produced at a scale of 500 g every three days (being that it takes three days for the nanowires to morph using the given method).

My director has asked me to scale this up by 10 to 5 kg daily and design a continuous, steady state process fit for commercial purposes.

The original challenge I ran into was the "continuous" aspect of the design. Given the 72 hour duration for the material chemistry to take place, it seemed highly difficult to somehow run the process continuously at steady-state rather than batch.

Upon further brainstorming, however, I determined that if I could get the reactants flowing through some sort of contraption that started at point A and got to point B in a three days length with enough turbulent flow to drive the reaction, then, indeed, a continuous, steady state process may be possible.

I decided upon the plug flow reactor, which seems to best along the lines of what I need to accomplish this. One of my curiosities has to do with how well PFR's work with solid-liquid mixtures, and also if these come with extensive lengths of pipe or tubing to conduct material for three days? Any ideas/suggestions would be appreciated.

 

[Chestermiller]

As part of a research project, I've been asked to design an industrial scale model for the synthesis of titania nanowires. These are currently being produced at a scale of 500 g every three days (being that it takes three days for the nanowires to morph using the given method).

My director has asked me to scale this up by 10 to 5 kg daily and design a continuous, steady state process fit for commercial purposes.

The original challenge I ran into was the "continuous" aspect of the design. Given the 72 hour duration for the material chemistry to take place, it seemed highly difficult to somehow run the process continuously at steady-state rather than batch.

Upon further brainstorming, however, I determined that if I could get the reactants flowing through some sort of contraption that started at point A and got to point B in a three days length with enough turbulent flow to drive the reaction, then, indeed, a continuous, steady state process may be possible.

I decided upon the plug flow reactor, which seems to best along the lines of what I need to accomplish this. One of my curiosities has to do with how well PFR's work with solid-liquid mixtures, and also if these come with extensive lengths of pipe or tubing to conduct material for three days? Any ideas/suggestions would be appreciated.

Why not use a CSTR? Then you could keep it well mixed and the solids properly suspended all the time. Of course, in that case, there would be a distribution of residence times. Would that be detrimental to product quality?

Chet

 

[JeweliaHeart]

I considered a CSTR; however the process could not be continuous if the material were sitting three days being stirred in a tank. The fact that PFRs usually come with a certain length of pipe so that there could be a continuous flow of material geared my decision in this direction.

It is important as you mentioned to keep it well mixed and the solids properly suspended all the time. I figured the turbulent flow through the PFR might be enough to accomplish this, perhaps?

 

[Chestermiller]

I considered a CSTR; however the process could not be continuous if the material were sitting three days being stirred in a tank. The fact that PFRs usually come with a certain length of pipe so that there could be a continuous flow of material geared my decision in this direction.

It is important as you mentioned to keep it well mixed and the solids properly suspended all the time. I figured the turbulent flow through the PFR might be enough to accomplish this, perhaps?

The letter C in CSTR refers to continuous. Flow is introduced continuously with an inlet stream, and is removed in an outlet stream. The volume of the tank divided by the flow rate is equal to the mean residence time.

Chet

 

[JeweliaHeart]

The letter C in CSTR refers to continuous. Flow is introduced continuously with an inlet stream, and is removed in an outlet stream. The volume of the tank divided by the flow rate is equal to the mean residence time.

Chet

My problem with the CSTR (Continuous Stirred Tank Reactor) is that, although it is being stirred continually, any new material flowing in is being mixed with the material that is already partially reacted. This is the problem with using "a tank" rather than a "pipe-like" system.

The material must be reacted for three days while being stirred continuously and any addition of new reactants during the 3 day residence time will result in lower product quality.

A CSTR would work just fine for a batch process, but seeing that this process is intended to be operated at steady state, I figured a PFR was a better choice.

 

[Chestermiller]

My problem with the CSTR (Continuous Stirred Tank Reactor) is that, although it is being stirred continually, any new material flowing in is being mixed with the material that is already partially reacted. This is the problem with using "a tank" rather than a "pipe-like" system.

The material must be reacted for three days while being stirred continuously and any addition of new reactants during the 3 day residence time will result in lower product quality.

This was the reason I asked the following question in my first post: "Would that be detrimental to product quality?"

A CSTR would work just fine for a batch process, but seeing that this process is intended to be operated at steady state, I figured a PFR was a better choice.

To guarantee good levitation of the solids in the reactor (as well as true plug flow without any Taylor dispersion), you could go to a static mixer, such as a Kenics Mixer, which uses twisted tape elements along the tube and results in a very uniform residence time distribution.

Chet

 

[JeweliaHeart]

This was the reason I asked the following question in my first post: "Would that be detrimental to product quality?"


To guarantee good levitation of the solids in the reactor (as well as true plug flow without any Taylor dispersion), you could go to a static mixer, such as a Kenics Mixer, which uses twisted tape elements along the tube and results in a very uniform residence time distribution.

Chet

Thanks for your help. It is nice to get feedback from someone who has more experience in the field than I do. Are you aware of any Kenics Mixers or PFRs with enough length to sustain a process that takes three days to react?

I estimated that at a feed flow of 1 ft/s, it would take approx. 49 miles of pipe length to sustain the 72 hour residence time. I realize this is a very high number, perhaps even to the point of not being a viable option for most firms.

Of course, however, these may in fact exist. If so, are you aware of any?