Classification of ore reserves based on geostatistical and economic parameters

"I would like to comment on the classification scheme of ore reserves based on geestatistical and economic parameters as proposed by Wober and Morgan in the CIM Bulletin, January 1993.In principle, a classification system that combines quantitative data for levels of confidence and errors with economic expectations must be welcomed. I would like to point out, however, two aspects of the proposed classification system which might create ambiguities, and therefore problems, by working with two levels of confidence and two kinds of block sizes. A normal classification scheme reflects the information density, the highest reserve category such as class A, or proven, or measured reserves having the highest density. This is not necessarily the case in the scheme of Wober and Morgan (1993), unless the definitions are amended.I. Assuming, as the authors do, that the estimation errors show an approximately normal clistribution, at the 95% confidence level, the standard deviation has to be multiplied by about 2 for the error and, at the 84% confidence level by about 104. Reserves (or resources) with the same or even higher information density than class A reserves, which at the 95% confidence level would not fulfill the required economic criteria at the threshold level, i.e. not qualify as ""ore"" at all, could be retained ""on paper"" to class B reserves by just multiplying the standard deviation by the factor 104 instead of 2, thereby narrowing the error.This statistical ambiguity was partly inherent in the old German classification scheme which used varying confidence levels and errors (GDMB 1956; Jahns 1959). In the new German scheme, which was later adopted in Austria in a moclified version, the level of confidence was therefore fixed at 90070 for all reserve categories (Wellmer 1983; ONORM 1989)."