| Summary: | An abundance of computerized exploration drillhole and production blasthole data from Cominco Resources' Buckhorn Mine in Nevada provides the opportunity for a comprehensive study to. evaluate various grade'estimation techniques. The Buckhorn ore-body is a bulk mineable, volcanic hosted, epithermal gold-silver deposit that formed in a Miocene hot springs system.
Forty separate block models, estimated from exploration drillhole data, were compared with 12 block models derived from blasthole data. The effect of choosing a smaller mining size block over the larger block size that might be dictated by drillhole spacing was examined in some cases. Some block models and datasets had geological constraints imposed on them while others did not. For each of four major block model
configurations various polygonal based, inverse distance and geostatistical estimates were calculated and compared to each other and to known production data. During the course of this study, many of the common statistical and geostatistical techniques for analysis of data were employed and in some cases, the methods themselves were questioned.
Results showed that estimating small, mining size blocks is important regardless of the grade interpolation method used because the best of these block models, although less accurate predictors of actual grade at any given location, tended to reflect the overall true grade distribution of the ore blocks which were eventually mined. Because all ore estimates are calculated from nearest exploration composites, some of the small block size models can be seen as fairly accurate, geometrically correct approximations (one possible realization) of the true, unknown shape of the orebody.
It was found that, in addition to separating oxide from sulfide exploration composites, the imposition of a second type of geological boundary, an ore zone outline within which ore was likely to occur and outside of which there was little basis for predicting the existence of recoverable reserves, measurably improved the estimates of tonnage and grade.
And finally, the method of calculating conditional probability proposed by G.F. Raymond (Raymond 1979, 1982, and 1984), and as further refined here, was found to be the most accurate, stable, and generally useful of the various methods used to generate block models and estimates of ore reserves at the Buckhorn Mine. === Science, Faculty of === Earth, Ocean and Atmospheric Sciences, Department of === Graduate
|
|---|
