Chemical reactors come in a variety of designs. Ordinarily, they range from bullet shaped pressure vessels to a pipe for plug flow reactions to a variety of cylindrical vessel designs. A big metal reaction vessel has several names- a pot, kettle, or reactor. Reactors can be customized with add-on components to suit specific requirements for agitation efficiency. Reactors can be used for continuous reaction as in the case of a CSTR, or for batch and semi-batch operations. Custom reactors may be built to provide unique performance specifications.
General purpose reactors can be purchased new or used. They come in a variety of materials of construction. Glassed reactors have a layer of vitreous glaze on the interior walls- often blue in color- and are resistant to corrosion, but may be harmed by thermal shock or electrostatic discharge.
Steel and stainless steel reactors come in a variety of alloy compositions. Hastelloy reactors can be acquired for enhanced resistance to corrosive materials, but at a steep price premium. Vessels with various types of cladding are available- Zr, Ni, Ti, Monel, Inconel, Hastelloy, Cupro Nickel. It is possible to obtain titanium or tantalum condensers for pots with particularly harsh duty.
Processes that require highly specialized materials of construction are usually more expensive. This can put considerable constraints on the process economics, since it is desirable to have the product requiring the specialized materials pay off the extra costs in a reasonable time period. This pay-off is in the form of a product price premium and/or depreciation. Taking on a project requiring specialized equipment often requires the cost analysis skills of an engineer to throw together a business case study. Perry’s Chemical Engineers Handbook is an excellent resource for this kind of activity.
Agitators are a very important part of the reaction vessel system. Motors, gear boxes, and impellers of various performance specs can be mixed and matched for projected requirements. Impellers are power absorbing implements. They absorb power from the drive motor. The job of an impeller is to dump the required number of watts per kilogram of solution into the reaction mixture to provide satisfactory shear. The energy required depends upon the geometry of the impeller and the density and viscosity of the mixture.
When trying to simulate a reaction on the bench top, it is critical to reproduce the big reactors shear at the smaller scale. Very often, this means that the rpm must be adjusted upwards to get the proper energy transfer. A great resource for this kind of work is the Pilot Plant Real Book, by Francis McConville.
Lamentations on Chemistry 
“Common sense” is linear. The real world has large exponents and serious punishments for ignoring them. Going from bench to pilot plant is an education.
Alright, finally I’ve got someone who can answer this question for me :
What is the ‘average’ scale-up I can do and still attain similar yields? 1X, 2X 10X???????
My assumption is that stirring and heat flow through the container are important factors. I’m also interested in what ‘types’ or reactions are more or less amenable to scale up. Additions (grignards) or nucleophilic substitutions, etc.Obviously if you’ve got an insoluble catalyst, this could be problematic, so assume we have soluble starting materials.
It’s a broad question, but any trends you’ve seen would be valuable.
Danke
I agree. Even a small reactor can accumulate a large amount of energy.
Hi Fritz,
I have seen small and large-scale ups: 2X and > 10X. I would beware of rules of thumb here. It depends on the process you’re scaling up and how much you understand the thermal behavior or the proclivity for a runaway. A process that accumulates unreacted reagents is one that requires great care in scale up. In fact, you should avoid the accumulation of energy all the time. A reaction that quits evolving heat when you stop the feed is the best situation to have. It is controllable.
The generation of non-condensable gases and the bp of the solvent system also figures in here. If your utilities cannot knock down gases that are evolved from the rxn, then you have to rethink the reaction scheme a bit. If your ability to control a reaction does not follow the scale, then you’re in for trouble.
These ramblings point out issues that must be considered before contemplating a scale-up. The magnitude of the jump depends greatly on how prepared you are to anticipate potential problems and eliminate hazardous conditions.
It is always more desirable to eliminate hazardous conditions than rely on better incident response.
Danke Herr Doktor!
Bitte.
Anybody out there care to comment on the relationship of all this to the current microscale craze in teaching organic chemistry and whether they have observed 1) a net decrease in explosions in sophmore orgo lab 2) a net increase in junior research lab explosions or 3) a real change in overall hazard to the only one who counts – the teachers?
Bill asks a great question. C’mon people. Cough up some feedback!