Optimising paper production

Faced with rising raw material and energy costs, pulp and paper producers are having to optimise the performance of their processes to remain competitive. Mike Spear reports on some of the latest advanced control techniques now being applied to these complex operations.

For a chemical or process engineer, the making of pulp and paper embraces just about every unit operation there is.

Starting with the mechanical handling and solids processing of the raw material itself — wood pulp or, increasingly, recycled fibre — the complex process moves inexorably through the pulping stage with its digesters, chemical treatments, washers and refiners, and on to the papermaking machinery itself, with its combination of wet-end slurry handling at one end and high-speed heat transfer at the other.

Not surprisingly, the pulp and paper industry was one of the first to appreciate the advantages of process control systems. In a typical large-scale pulp and paper mill, there could be anywhere between 30 000 and more than 100 000 I/O connections, mostly feeding through to the mill’s distributed control system (DCS).

This is conventional process control, analogous to DCS applications in other process industries such as oil refining and petrochemicals with equally large numbers of control loops.

But whereas those industries now also routinely use advanced control techniques, such as on-line process simulation and optimisation, to push their processes to the limit, the pulp and paper industry has lagged behind, largely because of that selfsame complexity.

Recent developments, however, both on the research front and commercially, could soon see the industry benefiting more from advanced control.

Under the aegis of the European intergovernmental Co-Operation in Scientific and Technical Research (COST) initiative, for example, teams of researchers across the EU are taking part in an action programme aimed at promoting modelling and simulation in the pulp and paper industry.

Starting last year and running through to 2008, the ‘COST E36’ programme has enlisted support from scientists and engineers in the industry, universities and institutions, and software suppliers.

According to the outline brief for the programme, no simulation model has yet been able to fully describe the processes taking place in the paper production process — particularly at the wet end of the paper machine, described as one of the most complex combinations of hydrodynamics and colloidal chemistry.

The programme will consider the use of dynamic process simulation, real-time simulation tools and model validation tools across this and the rest of the production process.

One of the first commercial attempts at modelling the complexities of an integrated mill can be found at Billerud’s Gruvön mill in Sweden. Here, ABB’s ‘Pulp & Paper On-Line Production Optimiser’ has been on trial for over a year, with early results pointing to an anticipated 3-5% increase in product quality and a reduction in the amount of make-up chemicals required.

The Gruvön mill produces around 600 000 tonnes of paper a year from six paper machines and three fibre lines. The pulp mill is supervised from six control rooms.

When Billerud recently invested in a new recovery boiler and evaporation plant to increase capacity and eliminate production constraints, the company decided to implement the ABB on-line optimiser to ensure that its anticipated increase in production actually happened.

Part of ABB’s IndustrialIT portfolio of control applications, the production optimiser runs alongside the DCS — or ABB’s latest variant, the 800xA Extended Automation system, introduced at the beginning of last year — predicting the dynamic behaviour of the system, considering all of the constraints and suggesting suitable remedial action when disturbances occur.

The on-line optimiser is the ‘missing link’ between advanced process control and a CPM (collaborative production management) system, according to Ulf Persson of ABB’s development team in Västerås, Sweden. By using a large number of on-line measurements across the production process, consisting of data such as production rates, buffer tank levels, laboratory data, capacity restrictions and planned maintenance stops, the optimiser can predict, in real time, future setpoints for all process sections.

As Persson says: ‘Integrated pulp and paper mills are complex systems. Consequently, it is difficult for operators to consider all variables involved simultaneously, predict the dynamic behaviour of the system and take appropriate action when disturbances occur. Therefore, the on-line production optimiser takes care of optimisation and determines the best operating strategy for the mill, regardless of whether it is currently in a situation of peace or upheaval.’

Results to date at the Grüvon mill show the success of the optimiser. Paper production is already up by 2-4%, there has been a 10-12% saving in chemicals, better decision support for production management, more efficient planning of maintenance, and less workload on operators when disturbances do occur.

Information based approach

While ABB’s production optimiser is based on rigorous modelling of all the production units in the mill, another approach to improving performance has brought significant savings at one of the world’s largest paper mills, Cenibra’s 940 000tpa plant in Belo Oriente, Brazil.

Here, an Exaquantum plant information management system (PIMS) from Yokogawa GMSC has helped achieve operational savings of over $2million since it was installed in 2004.

‘The company has been a user of Yokogawa distributed control systems for many years,’ says Nigel Bowden, md of Cowes-based Yokogawa GMSC, ‘initially with a Centum XL, now upgraded to the current Centum CS3000.’

When the Exaquantum PIMS was ordered by Cenibra it had a requirement for 7500 tags and 25 concurrent users, which has recently been expanded to 20 000 tags and 41 users.

The mill has reported three main areas where the system has benefited its operation — reduced downtime; reduced chlorine dioxide consumption; and reduced fuel consumption.

Before Exaquantum was installed, the plant suffered repeated blockages in the initial digester unit, which would bring virtually the whole line to a halt. Based on the data gathered by the PIMS and by cross-variable analysis, Cenibra was able to determine the root cause of the problem and reduce the level of blockages by a factor of six, resulting in additional production worth $60 000 a year.

Through historical data analysis from the PIMS, Cenibra also now has a better understanding of its use of ClO2 in the bleaching process. This has resulted in a 12.8% reduction in consumption, saving $1.2million a year. And Exaquantum has enabled Cenibra to quantify the energy balance involved in all its processes.

As a result the mill has saved 7500 tonnes of oil a month, equivalent to $570 000 a year.

Later last year a multi-disciplinary team from Cenibra developed a prototype Key Performance Indicator (KPI) system based on a standard Exaquantum/Excel access interface. This produces calculations for two key areas — loop control assessment and machinery assessment.

The success of the prototype has now led Cenibra to commission an intelligent KPI management application in an integrated web environment. This application can receive tags from Exaquantum and produce KPI calculations, the results of which are outputted to operators and engineering staff using its web graphics interface, and also fed back to the DCS to control production.