Prediction of Plant Process Performance Using Feed Characterisation — An Emerging Tool for Plant Design and Optimisation S P Suthers 1 , J M F Clout 2 and E Donskoi 3 1. MAusIMM, Metallurgist Iron Ore Processing, CSIRO Minerals, PO Box 883, Kenmore Qld 4069. E-mail: Steve.Suthers@csiro.au 2. Deputy Program Manager, Iron Ore Processing, CSIRO Minerals, PO Box 883, Kenmore Qld 4069. E-mail: John.Clout@csiro.au 3. Post Doctoral Fellow Iron Ore Processing, CSIRO Minerals, PO Box 883, Kenmore Qld 4069. E-mail: Eugene.Donskoi@csiro.au ABSTRACT The ability to design a beneficiation process for a new orebody based on particular feed characteristics is a powerful and practical tool. A new technique has been developed where beneficiation outcomes can be calculated by combining feed size-by-size chemical analysis, mineralogy, liberation and particle characteristics with a series of simple theoretical unit separation models rather than a simple mass balance approach or an exhaustive beneficiation test program. The new technique provides the ability to rapidly identify ore types that have the potential to meet product specifications whilst also rejecting others that will not meet specification even if a perfect separation was achieved. Beyond the broad scale decision making process, practical mineral separation curves from actual plant data can be used to more closely simulate the proposed process. This approach reduces the amount of laboratory and pilot plant work necessary by targeting what process stages will be required, increases early rejection of uneconomic options, identifies problematic ore types or unit operations and is a step towards estimation of final grade and recoveries achievable in potential circuit configurations. Examples are given for iron ore fines, where modelling predictions identified ores that would never reach target product grade, those that could be used for blending and those that were economic. The predicted outcomes of grade and recovery for a deposit composite are compared against actual test results using beneficiation processes such as grinding, desliming, magnetic separation and reverse flotation. A practical process configuration which achieved target iron grade and recovery was successfully determined during the testwork, guided by modelling results, whereas previous laboratory pilot plant trials were unsuccessful. The method can be equally applied to flow sheet development in plant design for heavy minerals sands, base and precious metals.