Abstract
This article aims to investigate the use of plurigaussian simulation as a practical tool to quantify geological uncertainty and internal dilution risk in a nickel laterite deposit. In laterite operations, sharp and highly variable contacts between weathering layers may directly affect selective mining, ore control, and the definition of ore and waste volumes. The proposed workflow integrates weathering surface modelling, unfolding, vertical proportion curves, lithotype rule construction and conditional plurigaussian simulation to reproduce the vertical organisation and lateral variability of the lateritic profile and to convert geological uncertainty into operational risk indicators. The methodology was applied to the Brejo Seco Deposit, a nickel laterite deposit developed over a mafic-ultramafic complex in Brazil. A database comprising drillholes and composite intervals was grouped into four main weathering layers: limonite, saprolite, boulder and bedrock. Ten equiprobable plurigaussian realisations were generated and validated through comparisons with original layer proportions, vertical proportion curves and vertical sections. The simulation results were then used to calculate node-based layer probabilities, generate accumulated two-dimensional risk maps, and estimate potential dilution volumes. The results show that unfolding improved the spatial continuity of the categorical variables and that plurigaussian simulation was effective at distinguishing areas where the deterministic geological model is strongly supported from sectors with higher contact uncertainty. For the limonite horizon, the estimated local dilution risk reached up to 90,000 t, demonstrating that deterministic modelling uncertainty may significantly impact mining performance. The proposed workflow provides a practical basis for identifying priority areas for model review, infill drilling, and geological control, and constitutes a useful decision-support tool for resource modelling and dilution risk management in nickel laterite mining.
Keywords
Get full access to this article
View all access options for this article.
