Forum on Advanced Process Control
Laurentian University, Canada, 5 June 2006

The recent International Conference on Mineral Process Modelling, Simulation and Control held on the 6-7th June at the Laurentian University in Sudbury, Ontario, was preceded by a one day Forum on Advanced Process Control. The Forum comprised a Panel of seven invited speakers. Each speaker gave a presentation and hosted an associated discussion, overall of some 40 minutes in length. The various topics involved both academic and industrial perspectives of aspects of the latest measurement and process control techniques and systems that are relevant to the Mineral Processing Industry. Subjects addressed included Economics, Image Analysis, Fuzzy Logic, Kalman Filtering, Optimisation, Model Predictive Control and Robust Control. The three main themes in the presentations were industrial application, products and theory.

In the morning, four presentations were given:

• Dr. Carl Duchesne of Laval University - “The Use of Multivariate Image Analysis and other Chemometrics Techniques in the Mineral Processing Industry”. The use of Principal Component Analysis to extract important information from images that relate to, for example, ore feed in grinding plants, froth in flotation and flame quality in kilns was described.
• Dr Fraser Forbes of the University of Alberta - “Estimating the Economic Benefits of Automation”. Examples were given to highlight pitfalls that can be encountered when assessing potential economic benefits if the inter-relationships between key factors are not accounted for.
• Dr. Eduardo Gallestey of ABB Schweiz AG - “APC and Scheduling Techniques for Real Time Economic Process Optimization of Plant Operations”. ABB’s “Industrial IT Solution Suite for the Minerals Industry” that incorporates new technology for mixed logic dynamic system optimisation was reviewed, with particular emphasis on Grinding Processes.
• Dr. Benoit Boulet of McGill University - “Robust Tunable Control of Mineral Processes”. This considered a basis of theory to establish a simple approach for tuning a Robust Controller.

There were a further three presentations in the afternoon:

• Dr. David Sandoz of the University of Manchester - “The Exploitation of Model Predictive Control in the Mineral Processing Industry”. Various applications made by Falconbridge Ltd to Nickel and Copper Smelting and Shaft Furnaces were reviewed and contrasted.
• Mr. Andrea Festa of SGS Minerals Services - “Expert Systems in Mineral Processing”. Aspects of the SGS approach to exploit Expert Systems and Fuzzy Logic were described – with particular reference to SAG Mill control.
• Dr. Andre Vien of Metso - “The Use of Soft-Sensors in Mineral Processing”. The application of a Kalman Filter based method of Soft Sensing was described in relation to a Pellet Plant Induration Furnace.

The audience, in all some 30 to 40 people, comprised a healthy mix of industrial engineers, instrument and technology suppliers and academics. Lively debate was a feature of the day – focusing on the contrast between the expert system methods for improved measurement and control versus the model based methods such as Kalman Filtering and Model Predictive Control and on the exciting prospects that Image Analysis offers for improving the ability to monitor and control effectively the minute by minute behaviour of Mineral Processing plants associated with Flotation and Grinding.

D.J Sandoz, University of Manchester, UK