1st MEETING OF THE PROCESS INTENSIFICATION NETWORK (PIN)

University of Newcastle upon Tyne

Thursday 15th April 1999.

MEETING MINUTES

 Background: PIN was formed on 1 January 1999, as a network activity supported, for the first three years, by EPSRC. The Network is based in the Department of Chemical and Process Engineering at Newcastle University, and its activities are directed with the assistance of a steering committee with industrial and academic representatives. The first full meeting was held in April 1999, when 64 members attended. The meeting comprised a number of scoping and technical presentations in the morning, followed after lunch by parallel workshops. This document forms the report on this meeting. A summary will also be given in the first issue of PIN News.

Welcome: Professor Colin Ramshaw, the meeting chairman, welcomed members to the University and to the first PIN meeting. He stressed that the aim was to make PIN self-supporting within two-three years.

Process Intensification - So What? Colin then proceeded to tell us why we should be interested in PI, in particular highlighting the principal drivers. These were seven in number, with capital costs reduction being at the top of his list. Showing that cost was a function of F x (main plant item), where F ranged from 4-8 and included civil engineering etc., PI could reduce F significantly. Intrinsic safety was the second feature attracting interest in PI, in particular as far as reduced fluid inventories were concerned. Energy efficiency - PI helps to reduce system driving forces, (temp., pressure & concentration differences can impair thermodynamic efficiencies). Colin cited electrochemistry and the Rotex chiller/heat pump as examples where energy efficiency arose out of PI.

The desk-top continuous processing system, for drugs/fine chemicals, would bring significant benefits - not just technological ones. Longer effective patent life, no need for plant scale-up, rapid response to the market and precise control of the process are all offered by such a PI concept. The example of this cited by Colin is the spinning disc reactor. For slower reactions, Colin took one of his popular 'biological' analogies - the 'intestinal' reactor for use when slower reactions are needed. Process control was the sixth driver - residence times can be reduced using PI from hours to seconds, therefore the concept of control is different. One may ask: "Do we need process control?"

Colin concluded by illustrating the breadth of application possibilities as the final driver for PI: It started in commodity chemicals, moving to fine chemicals/drugs, oil refining, waste treatment, food processing, metal recovery/refining, nuclear reprocessing and energy transformers.

Process Intensification Network - So What? David Reay, the PIN Co-ordinator, then detailed the reasons for the establishment of PIN, information on the steering committee, and how PIN can help members. He took some quotations from the questionnaire incorporated in the meeting reply form, highlighting the ways in which PIN could meet the needs of the individual members, and their organisations. David's overheads are attached to these minutes.

Process Intensification in Fine Chemicals: Francisco Bertran of SmithKline Beecham then explained why a company in the fine chemicals area was involved in PI. The main activities in SKB were batch processing and packaging of drugs. With sales of the order of 5-8 billion p.a., and 15-40% of current sales being from new drug applications approved in the last 5 years, R&D is critical and accounts for 12-20% of sales value.

Two years ago a development group, headed by Francisco, was established. The challenges to be met in the industry were, and remain, an increased flow of competing drugs, the market exclusivity time is shorter, and one needed to retain excellence in drug discovery and development, followed by commercialisation.

With regard to PI in pharmaceuticals, key factors were product cost, product quality and process reliability. There were also regulatory risks associated with safety, the environment and good manufacturing practices. Manufacturing investment was of course important, and the transfer of R&D to manufacturing technology needed to be done as quickly as possible. The advantages of PI were several: In batch processes flexibility was important (the spinning disc reactor could contribute towards this), and low volume processing was the norm. With regard to continuous processes, where SKB thought that most opportunities for PI existed, it could offer greater consistency, the process was easy to validate and control, a PI system had a lower capital cost, increased safety and the designs can match the reaction kinetics.

Francisco stated that PI would therefore deliver: Improved/consistent product quality; lower manufacturing investment; reliable and simplified technology transfer; high safety, and minimisation of the environmental impact. He illustrated in a most emphatic way the benefits that PI had brought to the production of a pharmaceuticals intermediate, using the spinning disc reactor: Reaction time reduced by 99.97%; plant volume reduced by 99.21%; impurity levels reduced by 93.33%, and reaction temperature reduced by 6.83%.

Process Intensification - An Oil & Gas Perspective: Brian Oswald of BP Amoco gave as his subtitle for the talk - more out for less in! Brian said that at the HEXAG meeting at BP (November 1998), Terry Lazenby, Chief Engineer, had likened PI to 'Honey, I shrank the plant!'. Brian saw it in the context of a process engineering 'time machine', PI giving leverage on any time-dependent parameter in a process - flux, gravity, throughput, hold-up etc. It was not just a replacement technology/unit operation, it was an alternative.

Brian took us through the range of operations which a company like BP Amoco undertook, from exploration to downstream conversions, transportation etc., most expressively put as 'rocks to customer'. The primary processes involved heat transfer and physics, and he cited 'liquefaction by chemistry' as a possibility. Oil and gas exploration/extraction and production was associated increasingly with smaller reservoirs, and increased flexibility in terms of operations in these areas was necessary. Brian gave an example of the changes dictated by exploration sites outside those in relatively shallow waters of the North Sea. West of the Shetlands, waters were 1000 m deep, while this became 2000 m and 3000 m respectively off West Africa and other regions. It therefore becomes necessary to to develop floating or sub-sea systems, as opposed to the North Sea platforms on rigid or tension legs. For these new concepts, PI technology is important. Health & safety issues are of course critical, with low inventories of systems being highly desirable.

Associated with extraction, and a technique for allowing the exploitation of additional reserves, is the concept of 'downhole processing'. One could produce high added-value products at the well-head. Unit operations here might include separations, reactions, biochemistry and energy production. Another concept might be the 'laboratory on a chip', for real time analysis of outputs/products, and control.

Transportation of products was not an obvious area for PI, but oil or gas in tanks and pipes represents a lot of residence time tied up doing nothing useful - transportation thus remains an untested area for PI (but not ruled out). The area can be linked to distributed processing and in distributed power generation.

At the refinery and petrochemical plants, chemistry begins to play a bigger role. With regard to refining, tightening of the product specification could be a benefit of PI. Products such as 'City' Diesel, 'reformulated' gasoline, low sulphur fuel oils and methanol for fuel cell-driven cars were all products demanding many expensive new process plants, with increased energy consumption in the processing stages, increased inventories etc. How can we improve things, and do better, asked Brian? Reduced pipework and lower installed costs, and heat integration were two answers. Both implied using PI. There was a strong incentive in petrochemicals to reduce capital expenditure (CAPEX) and operating costs (OPEX). While one could differentiate through enhanced product properties, there was also a need to integrate with upstream and refinery operations. A key to this was new reaction pathways. Brian cited distributed processing - the process on a chip, complementing his 'lab. on a chip'. In this context such a development could be used for screening catalysts.

Power generation was another area where PI could be employed. Brian cited the study for BP by David Reay & Colin Ramshaw on gas turbine reactors, with intercooler and recuperator reactors and the 'turbocracker' as being pertinent to this.

The general message which exemplified the approach of Brian's Group at BP Amoco was: (i) Do it differently - intensify, integrate, revisit the basics; (ii) Do it elsewhere - downhole, sub-sea, remote, in pipes, downstream; (iii) Do it less - challenge conventional specifications.

In conclusion, Brian said that on PI in the 'rocks to customer' business, the jury was still out on cost savings. However, there were no doubts concerning the value of energy benefits, inventories, safety etc. He saw PI as an 'enabling' tool rather than replacing existing technologies. But his final message to PIN was 'don't ever stop!'

Process Intensification in Combined Heat Transfer & Reactions: Andrew Green of BHR Group and Roshan Jachuck of Newcastle University did a 'double act' on heat exchanger/reactors. Andrew kicked off by asking the question: 'Why combine heat transfer and reactions?' He answered by saying that PI in this area was all about matching, and gave examples such as matching heat transfer to the rate of exothermicity of a reaction; matching flow behaviour to the reaction scheme; matching residence time to reaction time, etc. Andrew described the concept of heat exchanger/reactors (HEX/reactors) and the project which has brought together a number of manufacturers of compact heat exchangers whose products could be used for HEX/reactors to evaluate the units in this role.

He then went on to give examples, the first being aromatic nitration. The kinetics suggest a reaction time of <1 s is feasible, but the process is semi-batch and feed input can take over 12 h. Therefore there is a very low utilisation of capital. The producer could not meet the demand and additional by-products were formed. By-product formation could be significantly reduced using the HEX/reactor, and this was particularly effective when removed at source. An example of an ICI process which originally led to a 2% waste product generation was used to illustrate this - with heat removal effectively carried out at source (i.e. in a HEX/reactor) the waste was reduced to 0.01%!

Andrew revealed some of the first public views of the Chart Marston 'Marbond' reactor/heat exchanger. This, which could be made, for example, by chemically etching of plates followed by diffusion bonding, with very high surface/volume ratios (i.e. highly compact) was found to be better than the more established plate-fin heat exchanger with respect to minimising by-product formation. A 6" cube of 'Marbond' could contain sufficient catalysed surface to produce >100,000 tpa of a fine chemical!

Roshan Jachuck then took over to describe a second concept for combining reactions and heat transfer, the spinning disc reactor (SDR). This derives from the observation that rotating discs generate on their surfaces thin films with lots of surface waves when a liquid is fed on to the centre of the disc and flows outwards -and this is excellent for heat & mass transfer. One can achieve micro-mixing and fast reactions in such a manner. After describing the characteristics of the SDR (which has rotational speeds of typically 850 rpm), Roshan gave an example based on polystyrene. Here a batch gave a 30% conversion, and the SDR, with 4 discs but operating with a single pass, gave 80% conversion. Superior product quality was also achieved using the SDR. Polymerisation of polyester was also possible using the SDR. In the fine chemicals sector, it has been shown to give good shape and a tight size distribution when used to make barium sulphate crystals. Reaction times of milliseconds can be implemented when ultraviolet light is used to initiate polymerisation with the SDR.

In conclusion, Andrew and Roshan said that combining heat transfer and reactions showed significant benefits. The HEX/reactor and the SDR were complementary, application for each being different. Additionally, both technologies are available and are well understood.

Co-production of Power & Chemicals: There are a number of opportunities opened up if one brings together the techniques of process integration and process intensification. Frank Zhu of UMIST, (acknowledging his co-author Erik Tober of the University of Twente - a specialist in methanol processes), took as an example of a process where both techniques could possibly lead to benefits - the production of methanol and power.

Frank used as a starting point the ICI-LCM process, which involves, from a natural gas feed, steam reforming, partial oxidation reaction (with its heat being used to drive the reformer), and methanol synthesis, (the reactor heat here being used for feed preheating). He then asked: "Can we do better?" There were a number of options - one could use the syngas from the first reaction to produce methanol & power; one could make the overall process simpler; one could improve the material/economics of the overall process. One observation was that the H2 from the reformer and partial oxidation reactor was greater than that needed for the methanol. Therefore could the extra H2 be used for a fuel cell? Could the partial oxidation reaction be linked to a gas turbine and power produced as well? Both systems would still require a steam reformer.

The partial oxidation GT would involve delivery of syngas from the compressor, but the H2 does not satisfy the needs of the methanol process and one would have to purge a large quantity of the syngas. The direct oxidation of methane tends to have a low conversion rate (7%) and 50% selectivity.

Looking at the 'synergy of integration' involving partial oxidation, the PO GT, fuel cell and steam reformer, Frank that the approach was to build the 'superstructure' of the process - involving integrating all the possible different technologies. Then a process integration exercise would be carried out. Preliminary results, in terms of MJ power/kg methane, showed that direct oxidation to methanol would produce in the range 200-400 MJ/kg; fuel cells integrated in the methanol process would give 15-20 MJ/kg; while the PO GT would give 1 MJ/kg. The original ICI process gave 0.7 MJ/kg. The high power output case would, however, generate low product yields.

In order to progress the concepts, the challenges were, said Frank, development of a suitable fuel cell, development of a new catalyst for direct oxidation, and the development of a new gas turbine.

HiGee Applied to Heat Pumps: Bob Lorton of Interotex, the Cheltenham-based company managing the Rotex chiller/heat pump project, described the concept, and how PI had been applied to the unit - principally by using 'HiGee', the forces generated on rotating components. Rotex is a rotating absorption chiller/heat pump operating on a double effect cycle. Heat and mass transfer intensification is achieved by rotation (the unit having a diameter of 600 mm and a rotational speed of 800 rpm). Additionally, smaller 'compact' plate heat exchangers are used as solution heat exchangers within the body of Rotex.

Bob said the rotation allowed the generation of thin films of about 60 microns thickness. Coupled with the very small boundary layers and the highly energetic fluids resulting (75 g at the periphery of the rotating components), enhanced heat & mass transfer could be achieved. At the condenser, for example, a 'U' value of 65 kW/m2K was achieved and the film was only 10 microns thick. One of the problems in designing such a system was that oxygen and air were 'bad news'. Therefore an hermetic enclosure was necessary. Fluid circulation was achieved using a series of pitot pumps.

Currently a number of field trial units are being established, in particular in Northern Spain. The Rotex unit is housed in a 1.5 m high box for the trials. Bob concluded by saying that rotation provides an opportunity for unique fluid management and process intensification opportunities. Thin films result in markedly higher heat & mass transfer coefficients, and the system is cost-effective. In discussion, Bob said that the mass throughput in the absorption cycle in Rotex was 75 g/s, (for those wishing to calculate a chemical product output equivalent). The power requirement including all solution distribution etc. was 300 W, while the condenser delivered 7 kW heat. The Coefficient of Performance (COP) - the ratio of useful energy (hot or cold) out to energy in, was 0.95 with coolant at 40oC, providing 10 kW cooling duty - superior to any competing systems at present.

Oscillatory Baffle Reactors (OBRs): The OBR, stated Xiongwei Ni of Heriot-Watt University, was a combination of PI and product engineering (PE). The latter brought the characteristics of versatility and predictability to the concept. Process simulation was also a key issue in developing/exploiting the technology. The main groups working in the area in the UK were at Cambridge, Heriot-Watt and UMIST. At Cambridge the concept was being applied to paint dispersion. A network on oscillatory flow mixing (OFM) had been established, and there was strong activity in Canada (Malcolm Baird at McMaster Univ. on reciprocating columns) and in Australia (Tony Howe at Univ. Queensland on oscillatory baffle flow mixing).

Xiongwei described the concept of the OBR, stating that the oscillatory motion could be achieved in a number of ways, e.g. by bellows at the base of the column or oscillating baffles within it. Columns could be vertical or horizontal, continuous systems and straight or serpentine in form. An important parameter in design was the 'oscillatory Reynolds number', in addition to the net flow Re, which was the more traditional form. He said that OBRs offer, as well as enhanced heat & mass transfer, uniform mixing and particle suspension - analogous to a 'shaken', not a 'stirred' tank! In terms of PI in polymerisation, the use of an OBR reduces the number of waste particles, and also reduces both undersizing and oversizing of particles. Studies on three polymers showed that on average a stirred tank reactor (STR) gave product losses (out of spec. etc) equivalent to 2.5% of net sales value, while use of an OBR resulted only in costs worth 0.3% of net sales value. In yeast cultures, another application investigated, the OBR gave 75% greater mass transfers than in STRs.

Reactor volume savings compared to alternative methods were a factor of 8. One can carry out enhanced mixing and plug flow within the same device, and the OBR concept also offers much lower shear rates for similar levels of energy dissipation per unit mass than STRs. Low shear is of course particularly important for biological and bio-chemical processes. Batch, pilot and continuous processing are feasible. Batch processes include crystallisation for paracetamol and adipic acid. Any polymer reactions can also be accommodated in the OBR Polymer Centre.

Prior to the Workshop Sessions, Ming Tham of Newcastle University described the work he has been doing on the PIN Web Site. A sample page is appended. Ming showed us the first page: Welcome to the…..Process Intensification Network…..where size really matters! Pages cover subjects such as 'what is PI?', PIN, joining PIN, the PIN Newsletter, Members, Sponsors, the Steering Committee, a Diary, Projects, Abstracts, Links, Feedback form, etc….etc. The co-ordinator will inform all as soon as we are 'up and running'.

The Workshops

Four Workshops were run in parallel in the afternoon. These were as follows:

  • Barriers to PI and how to overcome them - Chaired by Mike Jones

  • Applications of PI to fine chemicals/drugs - Colin Ramshaw

  • Applications of PI to commodity/bulk chemical industry - Brian Oswald

  • Funding opportunities, including the EC FP5 - David Reay.

The chairmen reported back to all present after the workshops, and the notes below are based on their short presentations detailing the outcome of each workshop.

Overcoming the Barriers: Mike Jones summarised the outcome of his working group discussion. It was difficult to get industry to take step changes, and business needs to be convinced of the benefits as well (as academia). Also, how does one get people to work together? There is a need for multi-disciplinary teams in PI, but companies are wary of collaborations when one is working 'near market'. There is a need to define areas which need technology development. PIN could have a strong voice in this, both in the UK and overseas.

Mike identified communications/data base needs. Linking of PI with other techniques/tools, e.g. fast analysis. There was a need to identify areas for training, education, information etc. The problem is - no-one wants to be first! Pilot demonstrations could help here. Potential topics for the technology include: reaction + heat exchange + separations; logistics + process operation changes needed; process design + process control & transient reactions, process systems engineering for PI, process routes (e.g. LCA process).

Applications of PI to Fine Chemicals etc.: Colin Ramshaw said that the drivers were: Quicker to market, reduced capital cost, and a multi-purpose capability. The problem was to know how to overcome the cultural roadblock of addiction to batch stirred vessels. Having equipment available for demonstration on a client's site was important. It was also necessary to have a good dialogue between chemists and chemical engineers. A generic knowledge base is needed - possibly a book on PI?

The group believed that the way forward was to use demonstration equipment to convince management of benefits/get their attention, and to have an appropriate equipment supplier ready, willing and capable!

Applications of PI to Commodity/Bulk Chemicals: Brian Oswald stated firstly that the group agreed that the 'best' targets in bulk/commodity chemicals were very business-specific. Thus they decided to focus on the main drivers. These were seen as CAPEX/OPEX, energy/exergy, power, yield, inventory, the highest value product, and revamp/retrofit opportunities. Examples were seen as ethylene production, energy conversion and selective oxidation.

The group also looked at 'what next for PIN?' It was felt worthwhile to: (a) publicise the attributes of PI, covering sectors, types of process etc., (b) publicise case studies (including sensors and 'enabling' technologies), (c) encourage operators to understand current plant, (d) promote the integration of PI into current plant, and (e) get someone to do a real demonstration/pilot/revamp. Additionally, a list of all the 'players' was needed - industry, academics, vendors, 'funders', related technologies/enabling technologies etc.

Funding Opportunities: David Reay reported on this group. The possible sources of funding/collaboration available to date were: EPSRC - the EPSRC had set up the network with one aim being to help bring forward good proposals for support; DTI - the DTI was considering a LINK programme for later this year; the DETR, via ETSU at Harwell (Dr. Fiona Porter) - the DETR would fund R,D & D projects if it could e shown that there were energy efficiency benefits. The Energy Efficiency Best Practice Programme documents the specific programme requirements, and there is also a parallel Environmental Best Practice Programme where support opportunities for PI could exist. NOVEM in the Netherlands supports a Dutch PI Group (Prof. Van den Berg is co-ordinator) and member companies are keen to 'import' UK PI equipment for use in their plants.

The European Commission 5th Framework Programme was detailed in March 1999. Calls for proposals in both the Competitive and Sustainable Growth programme and the Energy, Environment & Sustainable Development programme have opportunities for PI-related projects. Safety and environment are motivators behind many of the detailed technical areas in the former programme, as well as better products, and lower costs.

Meeting Close: Colin Ramshaw thanked all participants and closed the meeting.