During the project kick-off meeting, called by “ICI” an HS1, the chemist described the raw materials, their properties, the reactions, the by-products, the waste streams, the optimum reaction temperature (for yield), the temperature below-which the reaction would not begin, and other things I can’t remember. It was a fine chemicals process for a **@@##** fungicide. The chemical engineer had outlined the processes and unit operations required and had concluded that an existing plant could be used, together with some other equipment from adjacent units, the idea being to make a few trial batches to enable a project costing and a quality/ acceptability evaluation. The process had been developed in their lab by the client, and had been repeated in “our” lab by the chemist.
The chemist was questioned by the safety engineer, who was running the kick-off. How exothermic was each of the processes in the multi-stage reaction, and how were the exotherms changed by yield and byproducts in prior processes? What were the accumulations of the various reactants fed-into the process at various stages in those reactions? (These were fed-batch processes carrying out what the Development Director called “elegant syntheses”.)
The chemist said that the processes were thermally neutral in the lab flask and no significant temperature rise was seen for any of the stages. He had deemed it not necessary to check the exotherms using the Adiabatic Calorimeter that the company possessed – nobody knew how to use it really.
This wasn’t deemed a sufficient answer and the timescale of the project required a quick answer. So, the safety engineer got the work done by a commercial lab. and the answers were not very re-assuring. One of the stages was exceptionally exothermic and another was very exothermic, so much so that the cooling capacity of the reactor chosen would not have been sufficient for the rates envisaged.
One of the stages was so minimally exothermic that the cooling capacity of the reactor would have quenched any reaction, (hence my statement earlier that a pumped-loop cooling system might be considered, often called tempered water).
In another example elsewhere, the chemist reported a reaction requiring control at 23.75 degC (for reasons of yield). The reaction didn’t occur at all below 21 degC, and other reactions started to predominate at 27 degC. The multipurpose reactor had jacket cooling and the chemical engineer asserted that he would attain the desired temperature by improved process control, there being no time or money to install a pumped-loop system.
The trials did work, but the yield was only 10%, with the process bouncing between “not reacting” and “other reactions predominating”.