Optimizing resource management under variable geological and mining conditions using a longwall advance model

Production planning in underground hard coal mines faces high uncertainty from geological variability. the longwall face advance is a key parameter determining production outcomes. this article models this advance rate based on local geological, hazard, technical, and organizational parameters. Instead of tonnage (a composite parameter), this research models the linear advance rate itself, representing the primary and most unpredictable component of excavation. this provides a utilitarian tool for decision-support systems and efficient deposit management. the research used integrated 5-year data from three hard coal mines, acquired from digital deposit models, scheduling systems, and operational reports, and aggregated monthly. Following a selection from 71 variables, a final set of 26 independent variables and one dependent variable (longwall advance per shift with production) was chosen. linear Mixed Models (lMMs) were applied to incorporate the hierarchical data structure (seams nested within mines). the model demonstrates a good fit, explaining 64% of total variance (conditional R2c = 0.64), while fixed effects alone account for 43% (marginal R2m = 0.43). Results indicate organizational factors have a dominant impact. the random effects analysis revealed 33.2% of residual variance stems from immeasurable, systematic differences between mines, highlighting the crucial role of mine-specific management factors. By successfully quantifying these diverse factors within a stable lMM, this study provides a model with improved predictive accuracy, establishing an effective foundation for operational planning and resource management.