Date of Award
Master of Science
Faculty of Communications, Health and Science
Associate Professor Lyn Bloom
The success of any mining operation greatly, if not entirely, depends on the accuracy of prediction of recoverable mining reserves. However, prior to mining, knowledge about the distribution of the Selective Mining Unit (SMU) is limited. The SMU represents the volume on which extraction of ore takes place and on which recoverable mining reserves are based. Realistic recoverable reserve estimates can be obtained from the grade-tonnage curve that corresponds to the unknown distribution of the SMU rather than to the distribution of exploration sample data. In general, if the reserve calculation, at the given cut-off grade, is based upon exploration drill samples, with much smaller support than the SMU, then there is a high probability of incorrect estimation of the tonnage and the grade of ore, and this can have serious implications for the economic side of the mining project. Various techniques for correction for the change of support of data, in other words change of the volume on which the data are defined, enable more accurate estimates of the distribution of the variable of interest (that is grade of a precious metal). The fact that the volume (support) represented by the variable is taken into account makes the estimates more reliable and, as we will show in the study, closer to reality. The distribution of the SMU is derived from the known distribution of samples by applying a correction model. Among these techniques arc two recent methods these arc a conditional simulation method detailed by I. Glacken and a kriging method due to A. Arik. This study aims to examine these two methods and compare them with the standard techniques. The methods will be applied to real data acquired from the Boddington Gold Mine in the south-west of Western Australia. In addition to accuracy, the practicality and simplicity of implementation of each method will also be discussed.
Janas, M. J. (2001). Change of support correction in mineral resource estimation. Retrieved from https://ro.ecu.edu.au/theses/1051