Abstract: |
For new mining projects, drilling data are invariably on a relatively large grid. The general indirect estimation technique of recoverable resources during these planning phases derives the unknown Selective Mining Unit (SMU) distribution estimates from the observed distribution of relative large kriged blocks(panels). The Gaussian model and the Uniform Conditioning(UC) technique provide an alternative consistent framework to achieve this task; this has been extended to the multivariate case(MUC). The drawback of the indirect methods is that only the probability distribution of the SMUs within local panels can be derived but not their individual spatial locations within the panel. Localised UC technique has been proposed to enhance the indirect UC by localising the results at the SMUs scale. The tonnages and metals represented by the grade tonnage curves estimated by indirect UC are decomposed and distributed into the SMUs within respective panels according to a ranking of the main element grade estimate of the SMUs. The local scale estimates for the other secondary metals, which are assumed to depend only on the main commodity grade are also derived, (the resultant local SMU estimates are referred to as Localised Multivariate Uniform Condition(LMUC) estimates). The MUC method provides a practical advantage, in that no specific hypothesis on the correlation between the respective secondary elements is required. The paper investigates the possibility of improving the LMUC estimates through multivariate block simulations which incorporate all the correlations of the secondary and main elements. As per the LMUC process, the tonnages and metals represented by the grade tonnage curves simulated through several multivariate simulations, are used to derive probable tonnages and metals, which are decomposed and distributed into the SMUs within respective panels (referred to as Localised Multivariate Simulated Estimates(LMSE). An additional advantage with the simulated results is that in addition to the LMSE localised estimates, the uncertainty of these estimates would be readily available from the simulated results, which will be useful for uncertainty modelling with regard to mine planning. The paper provides a brief review of the Multivariate Uniform Conditioning, Direct Block Simulation, the LMUC and LMSE techniques and presents a comparative case study based on a porphyry copper gold deposit in Peru.
|
