MINUTES OF THE 5TH PROCESS INTENSIFICATION NETWORK MEETING

HERIOT-WATT UNIVERSITY, 15 MAY 2001.

Colin Ramshaw, Chairman, (colin.ramshaw@ncl.ac.uk) opened the meeting, with 50 members attending, and introduced Bob Reuben, (R.L.Reuben@hw.ac.uk) Head of the Dept. of Mechanical & Chemical Engineering at Heriot-Watt, who welcomed members to the University. Bob outlined the role of the various campuses and said that the research & contract income was about 11 million annually, of which 32% was directly from industry. Of the 17.000 plus students, 32% studied on campus of which 45% read engineering. His Department had 400 undergraduates, of which 150 were chemical engineering students, with 60 researchers and 31 academic staff.

David Reay (DAReay@aol.com) then reviewed PIN activities since the last meeting (York in November 2000). Features included growth in membership to over 280, of which 50% are from industry. The PI Guide is timed to go out to consultation in July, with publication in December. A 'Summary Guide' will shortly be completed for publication in September. The next issue of PIN News will be published in late July - articles to David Reay before 10 July, please. EPSRC funding of PIN finishes at the end of this year, and discussions are under way to see if support from other quarters can be found.

PI at TNO: Jan Walpot of TNO in the Netherlands, (J.I.Walpot@mep.tno.nl) explained how TNO approached process intensification in the context of process 'innovation', which ranged from improvements to complete process redesign. Jan stated that PI as such had no implicit relationship to the degree of change. The main incentives for PI were product, economic and environmental advantages. He than ran through a number of case studies, the sponsors of the developments at TNO focusing mainly on fine chemicals, (9 sponsors including one cluster project). Examples which were not commercially confidential included the helix reactor, drying with zeolites/membranes, and the membrane slurry reactor.

Jan updated us on helix reactor developments, the reactor comprising coiled tubes, one of which carries the reactants while the other carries cooling water. Interestingly, the unit has a production capacity of the same order as a conventional unit, in 1/000 of the volume! It is also inherently safer, uses less energy, has a higher selectivity and therefore gives a better quality product. The tubes in the helix reactor were from a few mm to 5-6 cm in diameter, depending upon throughput and other variables.

Secondly, ceramic membranes have been studied under an EU project for esterification/condensation reaction, where one integrates reaction and separation. Zeolite is seen as an alternative in different processes, and this is put directly into the reaction mixtures to fulfil the separation. The maximum operating temperature could be a few hundred deg. C, this being a function of the membrane material. A third area of TNO activity was the use of superheated steam to improve drying. This gave better energy efficiency, and was faster. It was used for gypsum and ceramics drying and the pilot plant yielded evaporation rates of 100 kg/h, using steam at 100-400oC, and 2-5 bar(a) pressure.

During questions, Jan said that in the helix unit blockage could be a problem in certain cases. During the 4-5 week trials no problems had been met. One would take into account the diameter, turn radius and material type in assessing potential for blockage. First impressions were that it was very good with respect to blockage - liquids were the reactants, with different homogeneous catalysts. Colin Ramshaw pointed out that blockage by entrained particles was not too great in the plate PCHE-type reactor. He also queried why there was more interaction with the fine chemicals sector, finding the same himself. Jan felt that fine chem. Companies were much more focussed on product development, while TNO concentrated on the process - an area which was already an activity in commodity chemical companies.

The Gas Turbine Reactor: David Reay (DAReay@aol.com or D.A.Reay@hw.ac.uk ) then described the current feasibility study, funded by ETSU, on the gas turbine reactor - in which a gas turbine is being investigated for producing useful chemicals, as well as heat and power (overheads). The study involves 10 organisations, including Rolls-Royce, Norsk Hydro, ECN and BHR Group. The concept is based upon replacing a number of gt components with units which can undertake reactions. Possibilities include using exhaust heat with a recuperator/reactor, having reactions on one side of a reheater, carrying out endothermic reactions in turbine blades and intercoolers, and having a plate reactor carrying out combustion and reforming.

To carry the project forward to the R&D phase, further partners are being sought, and to this end a summary of the work to date will be sent to PIN members. In particular, David said that users, reaction specialists, CHE manufacturers and GT suppliers are needed.

Supercritical Fluids: Processing in supercritical fluids was the subject of a talk by Martyn Poliakoff of Nottingham University, (Martyn.Poliakoff@nottingham.ac.uk ). A web site related to the work at Nottingham exists - www.nottingham.ac.uk/supercritical/ Martyn explained what a supercritical (s/c) fluid is - one in which its temperature>critical T, and pressure>critical P. For CO2, a very common fluid used in s/c processing, this occurs at 31oC and 74 bar, while propane is s/c above 96oC and 43 bar. CO2 has power as a solvent similar to that of a light hydrocarbon, and solubility increases as pressure and density rise. A common use is coffee decaffination, and CO2 can be readily handled without major equipment problems.

The reactions being studied include hydrofomylation, acid-catalysed reactions and hydrogenation, the latter being the largest scale type. Usually an n-isomer is the desired product, and there is a problem separating the solvent from the catalyst. On an industrial scale one encounters mass transport limitations, and distillation destroys the catalyst. Taking acid-catalysed reactions, such as the synthesis of ethers, the conventional routes cause a lot of waste, and the acid catalysts themselves also generate waste. The process can also give unwanted byproducts. S/c CO2 is better than s/c propane for recovery in this process.

In the case of hydrogenation reactions, s/c CO2 overcomes the problem of the poor solubility of hydrogen in organics, which necessitates hard stirring to overcome the mass transfer limits.

La Roche first used s/c processes in pharmaceuticals in 1994. Chalmers Univ. + De Gussa carried out hydrogenation of fats and oils using s/c CO2 around the same time. In 1995 Nottingham worked with Thomas Swan & Co. trying to develop a more generic process capable of processing tonnes/annum in the lab., and one which was also clean. The reactor volume was 5 ml, production rate 1200 ml/h. One example of a product here would be the hydrogenation of isophorone, for the fragrance industry. Originally done at a rate of 7 ml/min, this has now been scaled up to 35 ml/min. The needs in Europe total 700 t/a.

In general, the s/c process is suitable for small reactors, giving high throughput, good safety, high conversion and high selectivity. The process is also environmentally clean.

A 30 litre reactor rig is now being built at Nottingham Univ.; meanwhile Thomas Swan will be bringing a 1000 t/a plant on stream in November 2001.

Reactive Distillation: Reactive distillation is seen as a key opportunity for PI, according to Hartmut Schoenmakers of BASF in Germany, (Hartmut.Schoenmakers@basf-ag.de). As an example, he quoted a process with ten unit operations (an extractor, reactor and eight distillation columns) which had been reduced to one - a reactor within a single distillation column. The production of methyl acetate at Eastman Kodak was such a case. Within BASF, in 1992 a large number of columns were used in similar processes, column number being much reduced by 1998, but targeted to be only one in 2005.

In this process, low pressure reactions can be combined with distillation. Several major companies studied in 1996-98 the processes most suitable for reactive distillation, and a new EU project, INTINT - Intelligent Column Internals for reactive distillation - is examining design and scale-up. This will be completed in March 2003.

Interestingly, process synthesis reveals that the yield of reaction in this context peaks at an intermediate reflux ratio, unlike normal distillation, where the higher the reflux ratio, the better. Simulators are currently not adequate for studying the process.

Scale up is a problem, associated with the 'separation of functions', as there is no technique at present for scaling up the combined unit operation of reaction and distillation. The problem is greatest with heterogeneous catalysts, homogeneous ones being not too bad. What is needed are experiments in mini-plants and the subsequent scale-up - but this is currently not possible.

When questioned about the application of the catalyst in the column, Hartmut said that it could be put in bags and suspended in the structured packing, or in bags on the distillation trays. A side reactor to the column gives better access to the fluid dynamics.

UV-LIGA Electroforming: The manufacture of micro-systems and components has implications for process intensification in several sectors, as well as in micromechanisms and electronics. Marc Desmulliez of Heriot-Watt University (M.Desmulliez@hw.ac.uk) described the UV-LIGA process and showed a number of examples of components and systems made using the technique. He began by describing the size range within 'micro' technology, which is in the range 100 nanometres to 1 cm. Below this, one is in the nanotechnology regime. Two examples relevant to heat and mass transfer were a micro-duct heat sink developed at MIT and capable of handling 1 kW/cm2 of heat removal, and the NASA SMART booster for the Space Shuttle, where the concept is to change one booster on this craft into millions of small boosters.

The UV-LIGA process was invented in Germany in the 1970's-80's. The basic steps are deposition, add a photo resist coating, transfer the pattern and develop the exposed resist. Then plate with metal and remove the rest of the resist to reveal the finished component - such as a micro-turbine. The master can be used for replication of many of the same, e.g. for injection moulding. Examples of products include micro-fans for cooling or as airflow sensors, heat exchanger micro-channels and freestanding structures such as car air bag sensors. Polymeric membranes for filters etc. have been made using photoresist methods, and Marc showed us a radial flow turbine of 3 mm diameter, actuate by the fluid, for driving a cutter used in medicine.

The advantages of UV-LIGA are low cost, a modest investment cost, and the fact that it is well-established. Drawbacks are the lower resolution achievable than with X-ray LIGA, and lower height-width (aspect ratios) achievable.

Marc described some features of the photoresist, such as columns of 330 microns height, a biochip with 3 micron channels, all featuring very smooth vertical walls, e.g. suitable as pads for micro-motors, and small gear wheels for clockworks. Further data are available on www.cee.hw.ac.uk/misec and www.cee.hw.ac.uk/microengineering

Short talks describing the laboratory facilities to be seen later in the day were given, including oscillatory baffle reactors (see www.COBRA.hw.ac.uk), the UV-LIGA facilities, bio-intensification work and heat pipes.

Impromptu Presentations

HSL/UMIST Facility: John Hare of the Health & Safety Laboratory, Buxton (John.Hare@hsl.gov.uk ) described the joint HSL/UMIST oil and gas testing facility which can be used to examine liquid-liquid and gas-liquid separations. After describing the plant and capabilities, John said that PI was one of the strategic areas for the facility, and HSL and UMIST were looking for projects, partners and customers to exploit the joint facilities now under development. Copies of the brochure on the facility an be obtained by emailing John, or viewing them on the PIN web site - the brochure needs Adobe Acrobat, while the talk at Heriot-Watt is available as a Powerpoint presentation. (full text and overheads)

The Flex Reactor: A new concept - the flex reactor - was described by Mark Wood of BHR Solutions (mwood@bhrgroup.co.uk). This is a reconfigurable flexible reactor in which different reactions can be examined. It is based on a shell and tube heat exchanger configuration, with in-line mixers in the tubes. A variety of end connectors - what would be U bends on normal tube bundles - incorporate facilities for sampling, dosing, separations etc., and these can be attached instead of plain bends as needed. The flexibility is typified by the fact that one unit could handle 80% of liquid phase reactions. The subject of two patent applications, the Flex Reactor is seen as a laboratory tool to test PI.

Kvaerner's PI Strategy: Mick Hilton (Mick.Hilton@kvaerner.com ) discussed ways in which the development of PI technology could be accelerated. He saw a brick wall between innovators and industrial/business users, and PI gets into what he called was a 'negative circle' which needed modular pilot plant with rapid access to allow people to test out ideas and address the risks. Concepts such as 'plug and play' skid mounted units for a range of reactor types were proposed. One example was current work with a company on a compact heat exchanger unit; another was loop reactor technology. Mick was keen to promote partnerships with those interested in process technology. (overheads)

Recent Dutch Activities: Henk van den Berg of the Dutch PI Group updated us on their activities (h.vandenberg@ct.utwente.nl) A PI Guide which may complement the UK version, concentrating on methodologies - 'How to do PI', is being prepared. Task forces exist on separation technologies and heat exchangers, and there is now a focus on the exchange of technology. (overheads)

Capillary Structures: Heat pipe wicks use capillary structures and these can handle high evaporative heat fluxes. Andreas Brautsch, a postgraduate at Heriot-Watt, (A.Brautsch@hw.ac.uk) describe some of the data he has already obtained on boiling in porous wick materials above flat or roughened surfaces. Such data are of use in estimating the capabilities of thermal control in micro-systems, for example. (overheads)

Delegates then went on a tour of several laboratory facilities. (overheads)

The next PIN meeting will be held on Tuesday, 27 November 2001 in Jesus College, Cambridge University, at the invitation of Dr. Adam Harvey and his colleagues in Chemical Engineering. Full data will be sent out in October.

++++++