Izvorni znanstveni članak
https://doi.org/10.17794/rgn.2026.5.2
PREDICTION OF MINERAL RESERVES BY A COMBINATION OF GEOPHYSICAL AND DRILLING DATA USING THE REGRESSION, COKRIGING, AND ARTIFICIAL NEURAL NETWORK IN THE MESGARAN COPPER DEPOSIT, IRAN
Kamran Mostafaei
; Department of Mining, Faculty of Engineering, University of Kurdistan, Sanandaj, Iran.
*
Hamidreza Ramazi
; Faculty of Mining, Amirkabir University of Technology, Tehran, Iran.
Mohammad Zamani Ahmad Mahmoudi
; Department of Drilling and Geo-engineering, Faculty of Drilling, Oil and Gas, AGH University of Science and Technology, Krakow, Poland.
Mitra Khalilidermani
; Department of Drilling and Geo-engineering, Faculty of Drilling, Oil and Gas, AGH University of Science and Technology, Krakow, Poland.
Dariusz Knez
; Department of Drilling and Geo-engineering, Faculty of Drilling, Oil and Gas, AGH University of Science and Technology, Krakow, Poland.
* Dopisni autor.
Sažetak
During a feasibility study of a mining project, a predictive model must be established to evaluate the profitability of the mineral deposit. In this research, three models were established by integrating geophysical data including Induced Polarization (IP) and Electrical Resistivity Tomography (ERT) with drilling data using the regression, Cokriging, and ANN techniques to determine whether the mineral production in the Mesgaran copper deposit is profitable or not. The Mesgaran deposit is located in the central part of the South Khorasan Province, Iran. The chargeability (M) and electrical resistivity (ER) data were used as input parameters, with copper grade as the output variable. The results of those predictive models were compared with an actual model obtained from twelve boreholes drilled after the feasibility study phase. Based on the conducted analyses, IP(M) data were found to be more suitable than ER data for constructing the 3D economic model of the copper deposit. In addition, all three predictive models estimated the mean copper grade nearly equal to 0.15%. Nevertheless, a significant discrepancy was revealed in their predicted tonnage and profit values for the deposit. Overall, we conclude that the ANN algorithm provides more accurate and reliable estimates for mineral resource evaluation. The results of this research can be successfully utilized in the financial analysis of mining projects.
Ključne riječi
Reserve Estimation; Induced Polarization (IP); Artificial Neural Network (ANN); Geostatistics
Hrčak ID:
352817
URI
Datum izdavanja:
27.10.2026.
Posjeta: 0 *