ABSTRACT
For the estimation of recoverable resources for mining projects when the data are limited, available techniques yield useful global estimates but all suffer from the problem of being unable to provide acceptable individual block estimates on a selective mining scale. For the purpose of an objective and practical comparison of available techniques an extensive data base for a 3D ore body was simulated. The data are on a close grid (13824 point values) to provide a basis for ‘actual’ block grades and for ‘actual’ grade, production and financial ‘life of mine’ profiles for feasibility studies. The position at the stage when only limited data are available, was covered by selecting a set of 216 widely spaced data points from the complete simulation. This data set was then used for block valuations based on ordinary kriging with an adequate search routine, direct conditioning applied to each block to remove the effects of smoothing and seven repeated simulations on a close grid. The block valuations for each technique provided comparative global grade, tonnage and relative profit estimates; also for twelve consecutive production periods and for the indicated grade and financial profiles for life of mine and feasibility studies. The results are analysed in detail in order to rate the relative efficiencies and uncertainties of the techniques.
