How details of the geometallurgical optimisation influence overall value

Document Type

Conference Proceeding


The Australasian Institute of Mining and Metallurgy

Place of Publication



School of Science




Originally published as: van den Boogaart, K G, Tolosana Delgado, R, Mueller, U and Matos Camacho, S, 2016. How details of the geometallurgical optimisation influence overall value. In Proceedings GeoMet 2016 Proceedings, (pp. 303–312). Carlton, Victoria: The Australasian Institute of Mining and Metallurgy. Available here.


The precise formulation of a geometallurgical optimisation problem is by necessity a simplification of reality. It includes implicit choices such as data availability or mining block geometry. Additional data can be orebody data, production data or market data. Often these choices are predetermined by the software, the workflow or the corporate culture, and as such are not even queried. There are several aspects that might play a role in the optimisation. They include not only the spatial model on which the selection of mining blocks and the potential mining sequence are based, but also the complexity of the processing model, the way in which uncertainty is treated, the scope of responsibility and the complexity of the model. In addition the timing of the optimisation and the availability (or lack thereof) of information will influence the outcome of any optimisation. We show that different modelling choices and different model formulations for the same mine can lead to very different processing choices, substantially different net present values (NPV), opposite investment decisions and so to overall different values for the entire operation. The selection of adequate simplifications and corresponding formulations for the optimisation problem is thus an important task in itself. For this purpose we introduce a method for quantifying the impact of certain simplifications prior to the optimisation. For example, the effect of mining block choices depends mainly on the spatial structure and the structure of the processing gain function. Both can be quantified beforehand. Based on the outcome, appropriate simplifications can be selected for the geometallurgical optimisation.