Driving the process

Kværner's process activities have roots stretching back to the origins of Britain's chemical industry.

Towards the end of the 19th century, Dr Ludwig Mond patented a process for the gasification of low-grade coal to generate 'producer gas', which was used both as a fuel and as a raw material for many early chemical processes. In 1901, he formed The Power-Gas Corporation to exploit this process, and Mond moved his research laboratory from London to Stockton. Over the last century, the company formed new links, and the usual vagaries of mergers, acquisitions and renamings has seen it under the banner of Davy Corporation, Davy McKee and Davy Process Technology. Most recently, its acquisition by Kværner saw it renamed as Kværner Process Technology (KPT). To accomplish this goal, KPT needed a larger state-of-the-art facility that would have the resources to fabricate and operate its own mini plants and follow through with engineered designs. After two years of planning and construction, a purpose-built new Technology Centre, costing £10million, was opened at Stockton-on-Tees in 1999.

Consisting of plant and laboratory facilities, chemical storage and office space spread over 4000m2 of floorspace, the centre is the home for the R&D activities of Kværner Process Technology, the technology and licensing arm of Kværner's Engineering and Construction (E&C) division.

The rationale behind much of the centre's work is to streamline the development of new processes. Traditionally, there are three stages necessary for developing and commercialising a large-scale manufacturing process - a laboratory stage, involving a range of different-sized small-scale rigs; a pilot-plant stage, around 5-10 per cent of commercial size; and the design, building and commissioning of the first full-scale unit.

Pilot plants are an essential part of this process, helping to validate the process and help with the design of a full-scale unit, as well as providing product for testing purposes. However, for a typical petrochemical process, just this stage is costly and time-consuming - a pilot plant generally costs in excess of £10million, plus operating costs, and takes over a year to build.

Kværner, like chemical companies such as BASF, is switching to a two-step process, replacing the laboratory and pilot-plant stages with a 'miniplant' stage, which have an output capacity of 0.5-1kg/hr. Generally, they cost around 10 per cent of a pilot plant, and can be built in six to nine months. Run continuously, for months at a time, using full automation techniques and arrays of monitors, they can cut as much as two years off the development process, Kværner claims.

However, miniplant techniques are not all plain sailing; the scale-up stage is much more challenging. The scale-up factor can be higher than 10 000, rather than ten or 20 typical for a pilot plant. However, Kværner says, this is 'a challenge which our engineers have met many times.' The most important information needed to design a reactor - and the instrumentation and control systems which operate it - are the thermodynamics and reaction kinetics which are taking place.

The behaviour of the fluids in the reactor is a particular problem, however. Custom-designed reactors can create fluid dynamics problems that cannot be replicated or modelled in the comparatively tiny miniplant reactors. Computerised fluid dynamics simulations can provide the key to solving the problems, but the uncertainty remains.

This has provided the impetus for one of the centre's collaborations. As part of a project with BP to develop a compact reformer unit for synthesis gas generation, the centre's researchers have built fluid dynamics simulation rigs to test combustion characteristics, the performance of custom-designed heat exchangers, and catalyst bed hydraulics.

The reformer is targeted for a gas-to-liquids plant which Kværner E&C is currently building for BP in Alaska, but BP and Kværner hold the patents jointly. The unit can be installed to provide a feedstock for a Fischer-Tropsch unit to produce long-chain paraffins, to be hydrocracked into a high-quality liquid fuel. Alternatively, the syngas can be used as a chemical feedstock in production of methanol, acetic acid, hydrogen, oxo alcohols, or direct reduction of iron ore.

Miniplant techniques at Stockton have also helped develop processes for hydrogenations, esterifications and complex separations, and development times have indeed come down. The most recent example, an ethyl acetate process, went from bench to commercial design in around two years.

Collaboration is an important part of the centre's work, both with plant operators and research institutions. The Stockton centre has close links with companies such as ICI-Synetix, Ineos, Sasol, BP, Dow-UCC and all the main catalyst manufacturers. The collaborations extend to co-funding the construction of miniplants.

Another success from these collaborations was the development of the world's first plant to convert ethanol to acetic acid. This began as almost a curiosity. While developing another process, KPT engineers noticed a side reaction where ethanol was dehydrogenated to acetaldehyde, and this reacted with more ethanol to form ethyl acetate. Realising that this could have commercial applications - ethyl acetate is normally made using acetic acid as a feedstock - they drew up a conceptual process flow diagram. This came to the notice of South African firm Sasol, which funded further research into the concept.

The team assigned to the project started by screening a range of catalysts for selectivity and conversion rates, and developing process flow schemes for reactor design and the purification of product. This was a particular challenge - the reaction generated a number of azeotropic mixtures which could not be separated by conventional distillation techniques. The team eventually designed a completely new distillation process to split the main azeotrope.

The second phase of the project, again at Stockton, saw the team run the reaction for 1000 hours to confirm that the chosen catalyst would not deactivate, and built a highly-instrumented miniplant to run the purification process continuously. The data generated from these tests meant that a commercial scale plant could be built without further experimental scale-ups and, after a series of cost estimates, Sasol awarded KPT the contract for a front-end engineering package to build the plant, along with a licence for the technology.

The plant was completed in April of this year, and began producing ethyl acetate in May. Within two days, it had exceeded its guaranteed capacity and was producing 99.99 per cent pure ethyl acetate, KPT says, which is slightly higher than most commercial ethyl acetate. The entire project, from initial concept to first commercial production, took five years - half the time of a typical plant design project.