Geology of bauxite deposits and their resource estimation practices

Abstract

This paper presents procedures and technical parameters used for resource estimation of bauxite deposits including the non-consolidated pisolitic bauxite (Weipa) and the intensely lithified types (Gove, Sangaredi, Az Zabira). Bauxite resources are usually estimated by drilling and sampling drill holes at 0·25 to 0·5 m intervals. Short sampling intervals are necessary for accurate estimation of the mineralisation contacts; this is despite the broad drilling grid which varies from 125 m centres for definition of the measured resources to 500 m centres used for inferred resources. Samples of non-consolidated bauxite are often beneficiated by sieving and removing the barren fine grained material before chemical assays. Consolidated bauxite ores are not beneficiated and processed in a conventional manner. However, in both cases, the overall precision error (CV%) of the samples, which is measured and monitored using field duplicates, does not exceed 10% (Al₂O₃ – 5%; SiO₂ – 10%, Fe₂O₃ – 7%, LOI – 10%). Bauxite density is preferably measured using the sand replacement method which is a formally certified technique for measuring bauxite density at the Australian deposits. However, a more recent approach is to use Sonic drilling to collect intact samples of the bauxite ore. Bauxite grade is estimated using conventional geostatistical techniques, most commonly by Ordinary Kriging. However, the method is applied after geometry of the bauxite seam is flattened using an equal thickness unfolding method, or, in some cases, using top flattening approach. Direct estimation of the bauxite grades without flattening their bodies produces incorrect estimates due to excessive smoothing of the estimated grades.