A Review of Numerical Models for Predicting Subsidence in Abandoned Mines (2015–2025)

Abstract

Subsidence in abandoned mines presents significant geotechnical and environmental risks, threatening infrastructure, ecosystems, and human safety, often decades after mining activities have ceased due to the gradual collapse or deformation of underground voids. This process can result in structural damage, soil degradation, altered groundwater flow, and sinkhole formation. The complexity of subsidence is controlled by the interaction of geological conditions, mining geometry, hydrogeological processes, and the mechanical behaviour of rock masses, which are often inadequately represented by empirical approaches. This review critically evaluates numerical modelling techniques used for subsidence prediction in abandoned mines, with particular emphasis on the Finite Element Method (FEM), Finite Difference Method (FDM), Discrete Element Method (DEM), coupled hydro-mechanical approaches, and monitoring-integrated techniques. FEM and FDM are widely applied for continuum-based stress–strain and deformation analysis, whereas DEM and hybrid FEM–DEM approaches are more suitable for modelling discontinuities, fracture propagation, and caving mechanisms. Coupled hydro-mechanical simulations enhance the understanding of groundwater-related deformation processes, while monitoring-based methods, integrating remote sensing, LiDAR, ERT, and InSAR, improve model calibration and validation. Advanced FEM–DEM models can simulate complex goaf behaviour with errors below 10%, while coupled hydro-mechanical simulations can predict surface uplift of up to 1.2 meters in water-affected areas, supporting more accurate analysis and sustainable management of abandoned mining regions. Studies from Albania demonstrate that integrating slope stability analysis, environmental assessment, and numerical modelling improves the understanding of mining-induced ground deformation. The review identifies major research gaps related to long-term validation, uncertainty in geological parameters, limited integration of monitoring data, and the absence of standardised modelling frameworks. Future research should prioritise hybrid multi-physics modelling, real-time monitoring integration, and standardised risk assessment methodologies to improve the reliability and sustainability of subsidence management in abandoned mining regions.