A real and viable alternative
From maximum to most efficient production using a continuous oscillatory baffled reactor

This article shows that the continuous oscillatory baffled reactor (OBR) manufacturing technology is real and viable in the production of a sophisticated and multistage organic chemical product. The product is made using hazardous and corrosive chemicals. The OBR offers uniform fluids mixing and excellent particle suspension, better mass/heat transfer than conventional reactor systems.


Nowadays we talk less of manufacturing technologies in chemical engineering as traditional batch stirred tanks have proved to work for a variety of chemical and biochemical engineering applications in both commodity and specialty sectors. In financial terms, the capital costs for manufacturing technologies are marginal to the overall operational profits, but the key question is always how much extra profit would any new manufacturing technology bring, in comparison with the existing one? The answer has to be “substantial” if such a new manufacturing technology would be adopted by industrial producers.
The next question is where any additional profits would come from if the new manufacturing technology is to be implemented? There are a variety of sources for additional profits, but the main one relies on consistent product quality, which entails uniform composition, size distribution and morphological properties. The consistent size distribution also means that there are less under and oversize particles in polymer applications, thereby a reduction in waste.

From the chemical engineering viewpoint, consistent product quality should automatically be associated with consistent fluid mechanical conditions in reactors, such as plug flow. We have learnt from the standard textbooks of Chemical Reaction Engineering that plug flow can be achieved by either running a number of continuous stirred tank reactor (CSTR) in series, or by operating high net flow Reynolds numbers in tubular reactors. To incorporate a number of CSTRs in series in a chemical factory would physically require a set-up of N CSTRs, each with an identical control unit. Ultimately this would significantly increase the capital costs and bring down the overall profits of the operation. For the tubular type of reactors, on the other hand, the main bulk phase would flow through the reactor with other reactants added in at different stages. Without going into too much detail, such an arrangement would certainly have many advantages over the former in terms of the inventory required and capital costs. The vital drawback for this type of reactor is, however, that significantly high net flow Reynolds numbers are needed in order for plug flow to be achieved, this translates as a very long reactor length for even low to moderate residence times. Consequently there have been very few continuous operations in relation to the production of fine and specialty chemicals. There are however examples of continuous operation in commodity chemicals, such as sulphonation and nitration on a scale of 100 to 1000s tons per annum.
The oscillatory baffled
reactor
Are there any other continuous reactors that allow both plug flow and much better performance than the two cases above? The answer is certainly yes. The oscillatory baffled reactor (OBR) would be one such reactor. The OBR generally consists of a column or tube containing periodically spaced orifice baffles with an oscillatory motion superimposed onto the flow. The mixing in an OBR is completely different from that in a stirred tank, and is provided by the generation and cessation of eddies. Such mixing is enhanced and much more uniform than traditional stirred tank vessels, as the intensity of the radial mixing is similar to that of the axial one. Each baffled cell effectively acts as a continuous stirred tank vessel. The use of an OBR to achieve plug flow condition coupled with a small steady flow component to obtain a long residence time allows production to be operated in a continuous manner. But the critical question is, has such reactors been proved to work in industries? Well, there were a few batch and continuous OBR units sold to various industrial companies worldwide during the 1990s, who attempted the OBR technology for manufacturing of different types of products and failed to make it into a commercial development! This article is to address this specific issue of continuous manufacturing utilizing the OBR with a real industrial and commercial development.
Process conversion
to OBR currently in hand
James Robinson Limited (JRL) is an autonomous subsidiary of Yule Catto plc based in Huddersfield. The company has a world-wide reputation as a manufacturer of hair dyes and intermediates, photographic chemicals, photochromic dyes and fuel marking dyes. An intermediate for a photographic chemical is currently produced on a several hundred-kilogram scale by a batch operation. The current process involves three stages: wetting, diazotization and cyclization, and two batch stirred tanks are employed and operated in a cascaded format, occupying a floor space of 12 m by 10 m by 10 m. The wetting and diazotization are done in the first stirred tank and the content of which is then discharged into the second one for the cyclization. The batch volume throughput after the diazotization is 3000 liters and after cyclization is 13 000 liters. Each batch operation takes about 18 hours to complete, including 12 hours for both the wetting and preparation of the buffer in the two stirred tanks; 2 hours for diazotization; ,1 hour for cyclization and 3 hours for isolation. The current operational parameters permit the manufacture of a single batch per day.
Converting the above batch process into a continuous operation requires one glass OBR unit of two sections (see picture): Section A consists of a number of
40 mm diameter tubes and Section B of
80 mm diameter tubes. The wetting and diazotization take place in Section A, while the cyclization in Section B. Parts of the glass tubes are jacketed for the purpose of cooling and heating. These tubes, each 2m long, are connected by U-bends and are arranged in a snake formation, starting from the top to bottom and then from the bottom back up to top and once more. The total flow path is of about 70 m in length, providing a residence time of about two hours. There are various ports fitted along the flow path for inputting components, sampling and measurements of temperature, pressure and pH.
Orifice baffles with an optimal restriction ratio are equally placed in both the straight sections and bends of the reactor. The baffles are made of PVDF and are designed to fit closely to the wall of the tubes. Each set of baffles is supported by three
3 mm diameter longitudinal PVDF rods. Two out-of-phase stainless steel bellows at the starting ends of the glass tubes provide fluid oscillation and are driven by a helical geared motor through a frequency inverter. A frequency range of 0-6 Hz can be achieved, and oscillation amplitude of 0 to 20 mm can be obtained by adjusting the off-centre positions of the crank in a flywheel.
For both sections, the net flow of the bulk aqueous phase is taken from the water mains and the flow rates are monitored by flow meters. Non-return valves are used in the flow inlets in order to reduce any propagation of oscillation upstream. In Section A, the limiting reactant is premixed with water and wetting agent and the mixture diazotized along the length of the OBR. At the same time, the preparation of a buffer for cyclization is taking place in Section B. The stages of wetting and diazotization are completed independently. Once the two streams meet, the cyclization stage commences and the product is continuously formed. The table summarizes the key operational indicators for both the existing batch and the continuous OBR processes.
Note that the yield and quality data are from batch OBR assessments. It is very clear from the table that by converting the existing batch production into continuous operation, we achieve product quality and yield consistent with the batch process; an increased productivity through the removal of down-time and more compact reactor set-up so that reduced capital cost, energy and space usage, improved intrinsic safety and reduced environmental impact become feasible. This has significant impact on the profit margins for such an operation. The project is scheduled to completion in nine months, and by the end of that, the OBR unit will be dismantled, transported and resembled at James Robinson Ltd and ready for continuous production on an industrial scale. The continuous manufacturing will replace the existing batch operation.
Applications in a range
of industrial operations
The continuous OBR technology is best suited to reactions with multi inputs of a multiphase nature, and can find applications in a range of industrial operations including phase separated synthesis, inorganic and organic crystals and particles, flocs, dyes, paints, pigments, surface coatings, adhesives, organic and pharmaceutical intermediates.
It is worth clarifying that the aim of this article is not to suggest that the OBR technology should replace or supersede conventional technologies, such as the stirred tank, but to highlight that the OBR is a real and viable alternative where appropriate, is easily accessible and can successfully be applied to niche processes or products, especially given that the modern trend of Chemical Engineering is to move away from large commodity manufacturing to specialty and high added value products. Continuous manufacturing is the key switching from the maximum to the most efficient production of those products.j
References
Mackley, M.R. and Ni, X., 1991, Mixing and dispersion in a baffled tube for steady laminar and pulsatile flow. Chem. Eng. Sci. 46: 3139-3151.
Mackley, M.R. and Ni, X., 1993, Experimental fluid dispersion measurements in periodic baffled tube arrays. Chem. Eng. Sci. 48: 3293-3305.
Ni, X., 1995, A study of fluid dispersion in oscillatory flow through a baffled tube. J. of Chemical Tech. & Biotechnology, 64: 165-174.
Ni, X., Cosgrove, J.A., Arnott, A.D., Greated, C.A. and Cumming, R.H., 2000, On the measurement of strain rate in an oscillatory baffled column using particle image velocimetry. Chem. Eng. Sci, 55 (16): 3195-3208.
Ni, X., Jian H. and Fitch, A.W., CFD modelling of flow patterns in an oscillatory baffled column, Chemical Engineering Science, 57, No. 14, 2002, 2849-2862.
Ni, X., Zhang, Y. and Mustafa, I., 1998, An investigation of droplet size and size distribution in methylmethacrylate suspensions in a batch oscillatory baffled reactor. Chem. Eng. Sci., 53 (No. 16): 2903-2919.
Ni, X., Zhang, Y. and Mustafa, I., 1999, Correction of polymer particle size with droplet size in suspension polymerisation of methylmethacrylate in a batch oscillatory baffled reactor. Chem. Eng. Sci., 54: 841-850.