Document Type : Original Research Paper


1 Department of Mining Engineering, Faculty of Engineering, Shahid Bahonar University of Kerman, Kerman, Iran

2 Faculty of Mining and Metallurgical Engineering, Amirkabir University of Technology, Tehran, Iran



At different stages of mining, we always face a degree of uncertainty. Some of these uncertainties, such as the amount of reserve and grade of the deposit, are due to the inherent changes in the deposit and directly affect the technical and economic indicators of the deposit. On the other hand, the heavy costs of the exploration sector often limit the amount of exploratory information, which necessitates the use of accurate estimation methods. In this work,we examines the modeling and estimation results using the conventional and simple kriging methods and the effects of the diverse indicators used in the classification of mineral storages or the parameters defining these indices. 127 exploratory boreholes with an average depth of 95 m are used to build the block model of the deposit in the Data Mine software. After the statistical studies, the 3D variographic studies are performed in order to identify the anisotropy of the region. A grade block model is constructed using the optimal variogram parameters.Then, using various methods to estimate the block model uncertainty including the kriging estimation variance, block error estimation, kriging efficiency and slope of regression, the mineral reserves are classified  according to the JORC standard code. Based on different cut-off grades, the tonnage and average grade are calculated and plotted. In this work, an innovative quantitative method based on the grade-number and grade-volume fractal model is used to indicate the classification of mineral reserves. The use of fractal patterns due to the amplitude of the variation is greater and more important than the standard and provides us with a better understanding of the deposit changes per block. The existence of a minimal difference between the use of the standard and fractal patterns in the slope of the regression method indicates less error and is a proof of more reliable results.