|Home News Conferences Commodities Publications Business Directory Resources Help|
:: Quest for Predictive Thickener Modelling
The latest phase of research to improve the performance of thickener technology in mineral processing uses advanced computational fluid dynamics (CFD) to develop a comprehensive predictive model for all zones within thickener operations. Thickeners are large tanks used in hydrometallurgy to concentrate the solids in slurries of tailings or valuable products.
According to Dr Phillip Fawell from CSIRO's Minerals Down Under Flagship, thickeners are often a neglected part of most hydrometallurgical circuits. "Even when thickeners only work reasonably well, people like to forget they exist and don't pay much attention to what goes into them," he says. "But when things go wrong they can become a major and very messy process bottleneck."
Dr Fawell is one of the key researchers in the current phase of the AMIRA P266 'improving thickener technology' research being undertaken through the Parker Cooperative Research Centre for Integrated Hydrometallurgy Solutions.
Past research has successfully used CFD to significantly improve the thickener feedwell process. Now, the use of CFD is being extended to the thickener zones below the feedwell where settling, consolidation and dewatering take place.
In their quest to create the first entirely predictive full thickener model, the P266 team is creating CFD models based on experiments across a range of scales, to better understand and predict how inputs impact on the settling system. One benefit of a predictive model would help avoid process bottlenecks, ensuring thickeners work with optimal efficiency and reduced cost. "A full thickener model would allow us to know in advance how a particular feed would behave through the entire thickener and therefore have the opportunity for better process control," Dr Fawell says. "Conditions could be tailored to a particular feed and optimal control strategies designed."
"The 'holy grail' for us is model-based thickener control, because if it's done properly you can then push your process harder, confident you can respond to any change in conditions." Currently, thickener operation is largely reactive, with adjustments made according to output measures sometimes made hours after the feed or flow conditions have changed. "To create a model to fully predict underflow density and handling properties, more needs to be known about sedimentation behaviour and what impacts upon it," Dr Fawell says.
In particular, researchers need to understand how key inputs, from the physical properties of the slurry itself to the application of flocculants and then raking of the beds, impact upon the density and rheology, or 'pumpability' of the final sediment.
Current experiments involve gamma-ray attenuation studies of selected settling systems in which a series of radioactive beams are used to measure solids concentrations as a function of time and settled height. These measurements are being used by mathematicians to create models describing the process that are then incorporated in CFD. "This dynamic relationship between experimentalists and mathematicians is a strength of the project," Dr Fawell says.
Concurrent research at the University of Melbourne into the impact of shear, or mixing intensity on aggregate structures within a bed also feed into the project. "We are currently developing a new research program to integrate our understanding of the mechanisms controlling thickener performance into a model that can be used for advanced control," Dr Fawell says. "There isn't anyone else in the world that can do what we do with thickeners in terms of modelling and measurement across all the thickener zones."
© 1998-2017, Minerals Engineering International