Modeling and ore reserve evaluation of Feyz-Abad Pb-Zn deposit using non-linear geostatistical methods

Extended abstract
Introduction
In the present research, geological-exploratory modeling of Feyz-Abad Pb-Zn deposit located in Isfahan province has been carried out as well as evaluation in terms of grade and amount of reserve. The exposed lithology units of the area with possible Paleozoic age from bottom to top, include chlorite schist, carbonate assemblages and green schists. Occurrence of zinc-lead mineralization with carbonate host in a metamorphic sedimentary volcanic sequence is one of the significant mineralization features in this region. The slightly altered limestone and dolomite, perform as host rocks of the deposit.
Materials and methods
The studied deposit area based on the topography of the region and amount and type of available exploratory information, has been divided into two distinct areas of Aghahesab and Kharzan mountain. So, modeling of the deposit and reserve evaluation of each region and the whole district have been undertaken separately. To know the statistical distribution of total Pb-Zn assay data and the correct choice of estimation method, first the initial statistical processing of the data was performed. Afterwards in order to fully understand and imagine the status of the deposit lithology, the log of all boreholes and drill holes, 3-D lithology model of the deposit and cross sections of lithology corresponding to a large number of the region designated profiles, were mapped. Depending on the mineral state, amount and type of carried out exploratory works and available exploratory information, among the variety of reserve estimation methods, accurate non-linear geostatistical methods containing ordinary log-kriging and indicator kriging have been employed. Then tonnage-grade curves have been plotted for cut off grades of 1.5, 2, 2.5 and 3 percent.
Discussion and results
The results of the research show that although both applied geostatistical methods are the most accurate of reserve estimation methods, however the difference between the results of estimation (grade and reserve) by means of the two methods is relatively high. The lack of sufficient exploratory information and the nature of the employed methods are the most important reasons for this subject but in general, the results of the indicator kriging method are more reliable. The amount of reserve calculated for various cut off grades in the study zones with the ordinary log kriging method is higher than the indicator kriging. This is due to higher blending of ore and gangue in ordinary log-kriging method than the other. For this reason, in general, the average grade for the variety of cut off grades in the indicator kriging method is higher than the ordinary log-kriging. The difference between the amount of reserve calculated by these two methods for the whole of Feyz-abad lead-zinc deposit for cut off grades of 1.5, 2, 2.5 and 3%, is equal to 26.9, 22, 27 and 17.8%, respectively.
Conclusion
Nonlinear geostatistical estimation methods have their own complexities, however, understanding the theoretical concepts of these methods properly and selecting their parameters correctly during estimation, will lead to very accurate, useful and desirable results. To increase the accuracy of ore reserve estimation of the studied area, it is suggested that the density of the minerals in various parts of the deposit to be determined by laboratory method with high accuracy. Then, through the index simulation method and determination of material type, an exact density is assigned to each block.