As mining operations evolve from open-pit to underground methods, the need for accurate and reliable resource classification becomes increasingly important. In a recent study conducted by聽糖心Vlog,聽we explored the application of conditional simulation techniques to assess geological uncertainty and optimize drill hole spacing in an Iron Oxide Copper Gold deposit with a goal of supporting more informed decision-making in resource estimation and mine planning during the first six years of underground operations.
This article is based on a presentation titled "Simulations Study to Classify Mineral Resources for the Underground Portion of an Open-Pit Copper Mine" delivered by Antonio Cortes, Principal Geostatistician, and Henry Kim, Principal Resource Geologist, at the聽9th International Conference on Geology and Mine Planning, Geomin-Mineplanning 2025 in Santiago, Chile.聽 Photos included throughout the article are courtesy of聽 Geomin-Mineplanning.
Importance of conditional simulation in mine planning
Conditional simulation is a powerful tool that enables geologists and engineers to generate multiple plausible realizations of geological and grade models. These realizations help quantify uncertainty and assess risk in various aspects of mine planning. For instance, conditional simulation can be used to evaluate recoverable resources by simulating different geological scenarios and estimating the proportion of ore above a given cutoff grade. It also supports reserve quantification by determining the probability of each block being classified as ore or waste.
In addition, conditional simulation is instrumental in modeling the variability of plant feed, which is essential for processing optimization. It aids in grade control by simulating dilution and ore loss, assuming one realization represents the 鈥渢rue鈥� grade. Furthermore, conditional simulation contributes to stochastic mine planning by incorporating uncertainty into optimization processes, resulting in more realistic forecasts. Finally, drill hole spacing analysis leverages conditional simulation to evaluate the uncertainty and cost-benefit of different drilling strategies, helping to identify areas that require more detailed information.
Application of simulation methodology in underground mining
We explored the drill hole spacing analysis use case with an Iron Oxide Copper Gold deposit that is currently being mined using open-pit methods but is transitioning to underground mining. The simulation process began with a thorough review of the geological database. Conditional simulations were then performed using Sequential Indicator Simulation for categorical variables and Sequential Gaussian Simulation for continuous variables. The simulations aimed to capture the spatial variability of geological features and copper grades.
聽To ensure accuracy, the simulations were validated through statistical comparisons and visual inspections. Fifty realizations were generated, each representing a possible 鈥渢ruth鈥� of the deposit. These realizations were merged by assigning copper grades based on geological classification: zero for Ocoite dykes, high-grade simulation values for high-grade shells, and low-grade simulation values for low-grade shells. The merged simulations were then reblocked to various sizes to align with expected planning and operational needs.
To evaluate the impact of drill hole spacing on resource classification, four synthetic drill grids were created, composite drill holes were sampled from the simulation results, and Ordinary Kriging was used to re-estimate copper grades for each grid. The study calculated F1 reconciliation factors for tonnes, grade, and metal across six production years (2026 to 2031). These factors were used to assess the accuracy of the estimates against the simulated 鈥渢ruths.鈥� Parker鈥檚 classification criteria were applied to determine whether the estimates met the thresholds for Indicated or Measured classification. Specifically, indicated classification required the estimate to be within 卤15% of the truth in 9 out of 10 yearly increments, while measured classification applied the same criteria to quarterly increments (approximated using 卤7.5% for yearly data).
Supporting more informed decision-making in mine planning
Highlighting the importance of drill spacing in resource classification, the results revealed that wider grid spacings introduce higher uncertainty and reduce the reliability of grade estimates, particularly in areas with high geological variability. Given that the yearly underground production is relatively small and highly variable, there is a significant risk that predicted production targets will not be met when using wide grid spacings.
This study demonstrates that conditional simulation is a valuable tool for resource classification in underground mining. By generating multiple realizations of geological and grade models, conditional simulation allows for a comprehensive assessment of uncertainty and supports more informed decision-making. The drill hole spacing analysis showed that tighter grids provide the highest classification confidence and are suitable for measured resources.
As the study showed, for Measured Resources, the required grid spacing is approximately 25x25 meters on a regular grid, and for Indicated Resources, the grid spacing should be at least around 50x50 meters on a centered grid. Wider grids, while more cost-effective, carry a higher risk of misclassification and may not meet the standards required for reliable mine planning.
Ultimately, the integration of conditional simulation into resource modeling workflows enhances the robustness of geological interpretations and supports the development of more resilient mining strategies.
With decades of experience in mineral exploration, resource and reserve estimation, and mining solutions, 糖心Vlog鈥檚 team of experts have strong experience in conducting studies involving conditional simulations, which help refine the delineation of mineral resource categories and support more robust mine planning. These techniques are well-established, widely accepted in the mining industry and fully compliant with international reporting codes.
Author
Glossary of terms
Ordinary kriging is a fundamental geostatistical interpolation method used to estimate grades or other continuous variables at unsampled locations, based on the spatial correlation structure defined by the variogram. It assumes a locally constant but unknown mean within a defined neighborhood, making it well-suited for mineral resource estimation where global trends are weak or absent. By minimizing estimation variance and honoring the spatial continuity of the data, ordinary kriging provides unbiased estimates that are widely used in block modeling and resource classification.聽
Sequential Gaussian Simulation (SGS) is a geostatistical method used to model continuous variables, such as grades or porosity, by generating multiple realizations that reproduce both the sample data and spatial variability. The process involves transforming the data to a normal (Gaussian) distribution, modeling the variogram, and then simulating values sequentially across a grid using kriging and random sampling from the conditional distribution. After simulation, the results are back-transformed to the original scale. SGS is widely used for uncertainty quantification in resource estimation and spatial modeling of continuous properties.聽
Sequential Indicator Simulation (SIS) is a geostatistical method used to model categorical variables by generating multiple realizations that reflect both sample data and spatial continuity. It transforms categories into binary indicators, models their spatial structure with variograms, and simulates values sequentially across a grid using indicator kriging and random sampling. SIS is especially useful for uncertainty quantification in geological domains with sharp boundaries, such as lithology or ore/waste classification.聽
Measured Resources refer to the most detailed and reliable assessment of mineral or resource deposits, based on comprehensive testing and data. They represent the highest level of geologic confidence and are used in late-stage technical reports and feasibility studies to determine if extraction is economically viable.聽
Indicated Resources refer to a category within mineral resource classification where the quantity, grade, and other characteristics of a deposit can be estimated with a moderate level of confidence, based on detailed and reliable exploration data. This level of confidence is sufficient to support mine planning and assess the economic viability of the deposit.聽
聽Stochastic mine planning is an advanced approach to mine planning that explicitly incorporates uncertainty into the process by using probabilistic models and simulations. Unlike traditional methods that rely on single-point estimates, stochastic planning considers a range of possible scenarios, accounting for fluctuations in factors like ore grade, commodity prices, and operational costs. This allows mining companies to develop more robust and adaptable mine plans that can better withstand the inherent uncertainties of mining operations.