Post Processing of SK Estimators and Simulations for Assessment of Recoverable Resources and Reserves for South African Gold Mines
This study is based on a comprehensive data base from a section of a large deep South African gold mine. The upper section of the area covered was accepted as providing the known data for purposes of estimating in a deeper extension of this section. Ore blocks in this extension were valued using the data from the upper ‘known’ area together with data in development raises typically 150m apart in the deeper extension area. Estimation techniques used were Simple Kriging (SK) with postprocessing, and Simulations and the recoverable block estimates were compared with the known follow-up ‘actual’ values of these blocks. The study shows that the direct SK post-processing and repeated simulation approaches, if applied efficiently, can provide equally useful tools for computing global recoverable resources. However, the direct SK post-processed technique provided the only advanced practical estimates of individual ore blocks for short-term mine planning, grade control and ore resource/reserve classification.
KeywordsOrdinary Kriging Sequential Gaussian Simulation Simple Kriging Kriged Estimate Dispersion Variance
Unable to display preview. Download preview PDF.