Uncertainty Assessment in Coal Resource Classification: a Comparison of Maximum Estimation Error and Superkriging Methodologies

Objective: This study evaluates the performance of two mineral resource classification methodologies, the Maximum Estimation Error (MEE) and Superkriging, to assess the uncertainty associated with thickness (m) in a coal deposit, considering annual and quarterly production volumes and different virtual sampling grids conditioned to the original data. Theoretical Framework: The research is based on geostatistical theories applied to resource estimation and classification, with emphasis on conditional simulation techniques and uncertainty quantification methods (MEE and Superkriging), aligned with CRIRSCO code best practices. Method: The workflow included exploratory data analysis, variography, normalization, and conditional simulation by turning bands. Virtual grids were extracted, and MEE and Superkriging criteria were applied for different production volumes. Results and Discussion: Superkriging proved to be simple and fast when applied to regular grids but was limited by not incorporating sample values, resulting in conservative classifications. MEE, although more complex, provided more consistent estimates, allowing the classification of resources as measured and indicated with 255 × 255 m grids for both quarterly and annual production. Research Implications: The absence of specific guidelines highlights the need to compare different geostatistical approaches across commodities, strengthening both academic knowledge and professional practice. Originality/Value: This study contributes to the literature by comparing MEE and Superkriging in homogeneous coal deposits, addressing a gap previously explored only in heterogeneous deposits.