Akademik Veri Yönetim Sistemi
International Geology Review, cilt.38, sa.9, ss.832-837, 1996 (Scopus)
Commonly employed ore-reserve estimation techniques, such as the polygon method and inverse distance weighting, use a weighted average of samples. The weighting coefficients are a function of the mining blocks in the deposit, but they shed no light on the variability of the orebody. Furthermore, these techniques do not allow a determination of the reliability of the estimates, but geostatistical estimation techniques allow a calculation of the measure of the error associated with the estimates. It is possible to find weighting coefficients for a given mining block and also data configuration that minimizes this error. This research describes a geostatistical evaluation of the Tulovasi borate deposit situated between the village of Osmanca and the Simav River in Bigadic, Balikesir (Turkey), The study begins with an analysis of composition; drillhole samples of variable length were aggregated to yield equal-length samples. Four hundred thirty-three composite samples show a mean of 26.67% B2O3 with a variance of 142.65 and a coefficient of variation of 0.45. Experimental variograms representing three main directions were calculated, and a simple spherical-type model was fitted to the experimental variograms. This was followed by three-dimensional block kriging. The deposit was divided into 40-× 40-× 6-m blocks, and the kriged estimate, together with the associated kriging variances for each block, was determined. Finally, the actual grade-tonnage curves were calculated. The grade-tonnage curves do not indicate where the individual blocks are located in the deposit; they simply summarize the numbers of blocks at specific grades. These improved estimates and the grade-tonnage curves can be used as a basis for mine-planning purposes. © 1996 Taylor & Francis Group, LLC.