Combined Use of Lithological and Grade Simulations for Risk Analysis in Iron Ore, Brazil

Abstract

The majority of economic feasibility studies of mineral projects use deterministic geological block models and grade estimation. Those deterministic block models do not allow project risk analysis. There is a smoothing of grade variability and the lithological dilution in each block is not considered. The uncertainty of geological block model can be evaluated by geostatistical simulation methods. The main objective of this project is to evaluate the impact of lithological and grade simulation in the economic studies of a world class iron ore deposit. An iron ore deposit from Carajás Province (Brazil), composed of low grade (jaspilite) and high grade ore (hematite), was selected as a case study. The hematite body is 9 km long, 3 km wide, and 300 m deep, with an average grade of 66 % iron. The uncertainty of the lithological contacts among hematite and waste/low grade rock is as important as the ore grade variability. For this study, drillhole database (hard data) and section interpretation (soft data) was used in order to improve the lithological conditioning. Different geostatistical simulations were combined to generate equiprobable realizations for lithologies (S L ), which are derived of superimposed different geological events; supergene and sedimentary/volcanic rocks. Two types of supergene events were simulated; the thickness of duricrust canga (2D TB simulation, S C ) and the transition of the weathering zone (3D SIS simulation, S S ). The levels of Banded Iron Formation and volcanic rock were simulated generating the primary rock facies (3D SIS simulation, S V ), independent from weathering simulations. The grade variables were simulated in cascade, one grade simulation for each lithology simulation. The results of the simulations were compared with the official reserve calculated with deterministic block model in order to measure the project risk and the impact of waste dilutions in the quality of ore product.