Stability Assessment in Underground Mining Systems

Ensuring the structural integrity of underground excavations is fundamental to safe and efficient mineral extraction. Stability assessment encompasses the evaluation of rock mass behaviour around openings such as headings, stopes and pillars, addressing potential failure modes including roof falls, pillar collapse and wall spalling. Assessments integrate empirical design charts, numerical simulations and data-driven models, informed by mechanical properties, in situ stress regimes, geological discontinuities and excavation geometry. Recent advances have refined finite-element and finite-difference methods, enabling high-resolution stress and deformation predictions, while machine learning techniques enhance risk profiling by accommodating uncertainty and complex parameter interactions. Monitoring systems—spanning remote sensing, geodetic surveys and instrumented supports—provide real-time feedback, supporting adaptive management. Such holistic approaches facilitate optimal support design, minimise unplanned dilution and maximise ore recovery, with applications across coal, metal and hard-rock mines worldwide. Cross-disciplinary integration of geotechnical engineering, computational methods and artificial intelligence underpins evolving best practice in underground mining stability.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Research from all publishers

Recent work has leveraged fuzzy logic and genetic optimisation to predict roof fall rates in coal mines, developing an inference system that integrates uncertainty in geological parameters with pattern-search techniques to fine-tune membership functions. This approach outperforms traditional statistical models, offering precise forecasts of roof instability with fewer rules. In hard-rock mining, gradient boosting decision trees (GBDT), XGBoost and LightGBM algorithms have been applied to pillar stability datasets, revealing that average pillar stress and pillar width-to-height ratio are dominant predictors. These ensemble models achieve accuracy exceeding 0.83, guiding pillar design and risk management. Numerical studies have also explored the influence of stope geometry on failure probability using Monte Carlo simulation coupled with finite-difference codes; factorial design and optimisation methods identify optimal ranges for stope dimensions, thereby reducing collapse risk. Together, these advances demonstrate the value of data-driven and numerical strategies in enhancing predictive capability and support design across diverse underground settings.

Stability Assessment in Underground Mining Systems publication trend

The graph below shows the total number of articles in stability assessment in underground mining systems across all publications each year (not limited to Nature Index journals).

Technical terms

Stope: Excavated cavity in a mine where ore is removed, bounded by walls, floor and hangingwall.

Pillar: Solid block of rock left between stopes or roadways to support overlying strata.

Fuzzy inference system: Computational framework that models imprecise or uncertain information using fuzzy logic rules.

Gradient boosting decision tree: Ensemble machine learning technique that builds sequential decision trees to reduce prediction error.

Finite-difference method: Numerical analysis approach that approximates differential equations by discretising the problem domain.

References

1. Fuzzy inference system using genetic algorithm and pattern search for predicting roof fall rate in underground coal mines. International Journal of Coal Science & Technology (2024).
2. Predicting Hard Rock Pillar Stability Using GBDT, XGBoost, and LightGBM Algorithms. Mathematics (2020).
3. Evaluation of the effect of geometrical parameters on stope probability of failure in the open stoping method using numerical modeling. International Journal of Mining Science and Technology (2019).