The Effect of Problem Formulation on Adaptive Processing Decisions

Geometallurgy is concerned with optimal mining and processing decisions based on a quantitative modeling of the ore micro fabric and the processes. The optimal choice is then inferred from a geostatistical model and a processing model (), typically by applying an optimization algorithm to select the best processing within a model formulation. The aim of our contribution is to discuss and show the effect of the problem formulation and on the performance of the optimal solution.

This analysis contains the effect limit input (e.g. grade vs. microfabric, joint distributions vs. correlation, ...), the scope of the decision (e.g. processing path, processing parameters, block decisions vs. shape decisions, decisions on sampling, blending, invest into capacity, sequencing) and of model accuracy (e.g. concerning unmodelled details).

The effects will be demonstrated with simulation examples in perfectly controlled stochastic settings. It can be shown, that problem formulation can have a major impact on the economic result, and that the effect of problem formulation depends on the actual setting.