How Ore Deposits can be Overestimated Through Computational Methods

Calculation of the amount of metal in an ore deposit or mining block should be approached by multiplication and integration of appropriate numerical manifolds describing the spatial distribution of the relevant variables, viz. grade, specific gravity and also thickness for bivariate data. Construction of a numerical manifold necessarily involves some form of interpolation but detailed characteristics of this are often not specified in various automatic methods, some of which even provide manifolds with discontinuities. A manifold used in calculations should describe in a geologically plausible way the spatial distribution of the variable of interest. Different approaches to the problem are discussed and compared. Particular computational inconsistencies emerge when "accumulation" is used as a variable in calculations of average grade. If an ore deposit is variable in relation to its sampling and especially if selection of mining blocks according to grade cut-off is a necessary operational decision, the biases introduced by accumulation methods can bring about serious problems of over-evaluation. Modern computational techniques make unnecessary the use of accumulations; because of potential inconsistencies it would seem unwise to use them in any ore reserve computation.