Science of Industrial Processes – Chemical Reactor Processes
November 30th, 2012
Preface
Quantifying chemical rate processes was the biggest challenge from the beginning of the 1960s for those companies wishing to control and optimise the very large (10,000’s tonnes per annum upwards) processes on which the chemical industry depends – hydrocarbon cracking and reforming, hydrocarbon chlorination, and polymerisation: ethylene, vinyl chloride, nylon and polyester chief among them, to which list was added polypropylene in the 1970s. This is not to say that challenges did not exist elsewhere – they did – but thanks to work at MIT in the 1930s and 40s, by the 1960s the behaviour of the other key operation dependent on physical chemistry – distillation – was by comparison with the knowledge of the key reactions, pretty well known and quantified.
The essential problems in understanding and quantifying chemical rate processes boil down to two: (1) how can you measure the products of a reaction and (2) how do you know under what specific concentrations of chemical reactants and temperatures, the measured products were formed. This problem has its exact counterpart in understanding polymer chain kinetics and fibre orientation kinetics as in the synthetic fibres and SAFIRE technologies. Both problems are micro-versions of the general manufacturing consistency objective: how do you discover where an inconsistency-causing fault enters the flow upstream of the point you are examining (usually the final quality control station).
At the level of 0.1% the gas-liquid chromatograph (GLC) attached to the outlet stream solved the first (measurement) problem completely for gaseous petrochemical species from about 1965 on. Since then sensitivities of the detection systems attached to the GLCs have been greatly increased and their range extended (e.g. with online mass spectroscopy in addition to hot-wire conduction detectors).
The second problem was completely solved for all the major industrial petro-chemical processes by the continuous flow stirred reactor. The basic principle is that if reaction is confined to a particular volume of space (or cell) where the gases or liquids are sufficiently well-mixed, then all the reaction will occur at one set of compositions and temperatures, namely those measured at the outlet of the cell. This applies both to reaction sets which are mainly catalytic and to free radical reactions where surfaces of the cell play a key part in the initiation and termination mechanisms.
Achieving the requisite degree of mixedness for this assumption to be valid depends on the flow rate into the cell, flow induced in the cell itself, and the speed of chemical reaction in relation to the induced flow. Broadly-speaking, the reaction rates of the principal measurable constituents of the major gas phase processes are in the order of 10-20% per second at the temperatures and pressures of industrial interest: 400-1000 degrees C; 3-30 bar pressures.
Two types of reactor were designed for this duty at ICI Corporate Laboratory, Bozedown: a 200 ml stainless steel reactor with a 100 ml basket carrying catalyst (particularly for the hydrocarbon reforming and ammonia reactions) and a series of silica jet-stirred reactors (JSR) [see Moulton medal paper] ranging in size from 25-300 ml, capable of withstanding 1,000 degrees C – particularly for hydrocarbon cracking, chlorination, and oxidation (see The Experimental and Computational Determination of Complex Chemical Kinetics Mechanisms, proceedings of the Royal Society, 1976). While the focus in the 1970s was on gas phase reactions, the JSR is equally applicable to biochemical liquid phase reactions, even where only tiny continuous reactant flows are available, the cell volume being reduced accordingly [see The Application of Jets to Chemical Engineering Operations].