Deep underground hardrock mines face escalating rockburst risk as depth and stress increase. This thesis develops and demonstrates an integrated framework for stress management and rockburst mitigation that combines destress blasting and destress drilling with energy‐based indices, advanced monitoring, and geostatistics, tailored to Swedish deep mining conditions. Four objectives structure the work: (i) derive a design framework for destress blasting from six decades of international and Swedish practice; (ii) construct a quantitative evaluation methodology that integrates the strain energy storage coefficient (F), brittle shear ratio (BSR) and burst potential index (BPI) with fracture and seismic observations; (iii) execute and analyse a controlled field trial of destress drilling at Zinkgruvan mine; and (iv) develop a geostatistical concept (semi-variograms and kriging) to predict destress efficiency at unsampled locations with quantified uncertainty. 

The methodology adopted for the thesis study integrates structured literature and case-history analysis, numerical and energy-based reasoning, and in situ experimentation with high-resolution 3D laser scanning and cloud-to-cloud (C2C) analysis. The destress blasting component organises key rockmass, stress, and explosive parameters into a conceptual decision framework and design guidance; the evaluation framework specifies how F, BSR and BPI are computed and interpreted alongside monitored fracture and seismic responses; the Zinkgruvan mine practical field trial isolates the mechanical effect of uncharged inclined boreholes by keeping production-blast variables constant and quantifying geometric outcomes through C2C and volume-added metrics; and the geostatistical study shows how semi-variogram modelling and ordinary kriging can map performance indicators and their uncertainty to support risk-aware planning. 

Across case histories and supporting analyses, destress blasting is shown to be effective but highly localised and transient: stress relief typically extends only a few metres from the blast, and benefits decay rapidly as faces advance, necessitating continuous inclusion of destress features in each round within burst-prone zones. Mechanistic interpretation links reductions in boundary tangential stress and strain energy density to blast-induced fracture networks whose extent depends on rockmass brittleness and charging/timing choices; highly brittle rocks are both more burst-prone and more responsive when patterns and charge intensities are matched to site conditions. 

The practical Zinkgruvan mine field trial (depth of 1285 m) provides quantitative evidence that destress drilling stabilises development drifts. Rounds with 46 mm, 4 m destress drilling holes inclined at 20° (roof and shoulders) exhibited up to 2.5 m3 less scaled ‘volume added’ per metre of advance and a 20 – 30 % reduction in C2C profile standard deviation relative to non-destressed rounds, indicating lower overbreak and improved excavation profile control. It was observed that the first two rounds after a destressed round also performed comparably well, evidencing a short-range residual benefit that dissipates by the third non-destressed round, an operationally important finding for sequencing and cost-risk optimisation. 

The thesis study advances practice by: (i) organising destress blasting design considerations into a transferable, Swedish-context-aware framework; (ii) unifying energy indices (F, BSR, BPI) with fracture/deformation and seismic monitoring for quantitative evaluation at excavation scale; (iii) providing the first high-fidelity, field-validated C2C/volume-based assessment of destress drilling in a deep, burst-prone European mine; and (iv) introducing a geostatistical prediction concept that generates mine-wide efficiency maps with confidence bounds to reduce hazardous measurement campaigns and guide targeted data acquisition. 

The main conclusions and recommendations are that destress measures must be engineered and applied continuously in high-risk zones; design should be matched to rockmass brittleness and in situ stress; evaluation should jointly track energy indices, deformation/fracture, and seismicity; soft-scaling (where appropriate) practices should be integrated to minimise added volume; and geostatistical mapping and digital tools (3D scanning/C2C, IIoT) should underpin adaptive, feedback-driven planning.

Stope instability remains a persistent and hazardous challenge in underground mining, impacting safety, efficiency, and sustainability. Traditional stability assessment methods, while valuable, are often limited by site-specific calibration, simplifications, and adaptability issues in dynamic underground conditions. While machine learning shows potential for improved accuracy, a critical gap persists in understanding how geotechnical factors interact in practice. This study introduces a novel, practical machine learning framework (Scikit-Learn) to predict stope instability, and crucially, to quantify the nuanced, non-linear influence and interaction of critical geotechnical factors in a shallow gold mine. Comprehensive geotechnical investigation (observations, lab tests, rock mass classifications, blast damage assessments) and advanced data analysis (Random Forest feature importance, RFE, decision boundary analysis) identified water ingress, blast-induced damage, and rock mass quality (RMR) as the most significant instability factors. Water ingress profoundly impacted stability, with moderate blast damage exacerbating instability under high water ingress. Rock strength showed comparatively lower significance. The developed model achieved robust predictive performance (accuracy: 0.83, precision: 0.88, recall: 0.83, F1-score: 0.83). Based on these insights, tailored support patterns (e.g. 22mm/16mm cone bolts, timber props) are proposed to mitigate specific risks. This research significantly advances targeted rock mechanics solutions by providing a deeper, quantifiable understanding of complex instability mechanisms, enhancing mine safety and operational efficiency in shallow gold mining. 

This study advances pillar stress prediction and design optimisation in hardrock platinum mining on the Great Dyke of Zimbabwe using advanced machine learning (ML) techniques, addressing significant gaps in traditional methods. Utilising Gradient Boosting Machine (GBM), XGBoost, NGBoost, Random Forest, and AdaBoost, the research evaluated a dataset of 503 observed practical insitu pillars, incorporating key features such as Depth Below Surface (DBS), Actual Panel Width, and Actual Extraction Ratio (AER). GBM and XGBoost emerged as top performers, achieving R2 scores of 99.58% and 99.44%, respectively, with GBM exhibiting an MSE of 0.3094 and RMSE of 0.5563. NGBoost added value with predictive uncertainty, enhancing risk management frameworks. The study also highlights feature importance, emphasising DBS, AER, and Actual Pillar Area as critical predictors, ensuring robust and site-specific design solutions. Practical outcomes include a 15% reduction in material overdesign and a 20% improvement in identifying high-risk pillars, contributing to safer and more efficient operations. Integration with real-time monitoring systems enabled dynamic adjustments, reducing pillar failure risks by 30% under evolving conditions. This research, the first of its kind on the Great Dyke, demonstrates the transformative potential of ML in mining engineering, providing a framework for safer, economically viable, and sustainable operations. This study paves the way for leveraging ML to tackle complex geological and geotechnical challenges in global mining projects by addressing predictive accuracy and uncertainty.

This research aims to optimize ground support systems for underground tunnels in geologically challenging environments, specifically addressing the reduction of Fall of Ground (FOG) incidents in a gold mine in Mashava, Zimbabwe. The study integrates advanced detection and classification methodologies to enhance tunnel stability and safety. Tunnel Reflection Tomography (TRT) was employed to identify unfavorable geological structures ahead of excavation, while core logging at 20 locations on level 7 provided rock mass quality assessments using three classification systems: Bieniawski’s Rock Mass Rating (RMR), Laubscher’s Mining Rock Mass Rating (MRMR), and Barton’s Q-system. The results consistently indicated poor rock mass quality, informing the design and refinement of a robust ground support system. Fallout height data from past FOG incidents and probabilistic key block analysis using J-Block software further validated the support system's effectiveness. The findings significantly reduce collapse risks and downtime, enhancing operational safety and efficiency. This research contributes to developing practical strategies and tools for improving tunnel stability in complex geological settings, offering valuable insights for future advancements in mining support technologies. The study's necessity stems from the industry's growing demand for innovative solutions to enhance tunnel stability in adverse geological settings, particularly in regions with limited access to advanced technologies or methodologies.

As mining progresses to greater depths, the challenges of high stress become more pronounced, often resulting in rockbursts that significantly impact deep underground mining operations. To address these challenges, Zinkgruvan mine in Sweden is testing destress drilling as a proactive measure to reduce the propensity for rockbursts and enhance the long-term stability of the mining drift, particularly in the roof and shoulders. Destress drilling holes, in this study, were drilled at 20° inclination on the periphery of the exploration drift and strategically placed ahead of development blasts. Laser scans of the drift were conducted before and after scaling, and the point cloud data was analysed using Cloud Compare software, with the Cloud-to-Cloud (C2C) algorithm employed to detect profile changes. This allowed for a comparison between blast rounds with and without destress drilling to assess the technique’s effectiveness. Results demonstrated that destress drilling reduced stress concentrations in the surrounding rockmass, as evidenced by reduced profile change. Blast rounds with destress drilling had up to 2.5 m3 less volume added per metre. C2C analysis showed 20 % to 30 % lower standard deviation and consistently lower mean deviation, indicating improved profile uniformity. These findings highlight the technical and operational benefits of destress drilling. 

Reliable stability assessment requires an objective and precise assessment of the rock mass quality classification. A deep learning model is developed to create a tool that can provide a rapid and precise assessment of the quality of rock masses. While there are empirical equations to determine RMR values from Q parameters, this study provides an advanced highly accurate deep learning approach to determine RMR values from Q parameters. This serves to reduce the amount of fieldwork related to collecting the rockmass data needed to independently assess rockmass quality using the RMR system and the Q system separately. The RMR values, like Q values, were first determined independently in the field. The deep learning approach was later used to predict the field-determined RMR values from the field-determined Q parameters. This means that each practical field measurement point had an RMR, and a Q value independently determined for it before the deep learning approach was applied. The six rockmass parameters of the Q system (RQD, Jn, Jr, Ja, Jw, SRF) are used as input in this model while the RMR is used as the output variable. In this study, the dataset contains 356 samples, 70%, 15% and 15% of the entire sample data are used to train, test, and validate the model, respectively. The predictive performance of the models was evaluated and compared using metrics such as R2, MAE, and RMSE among many others. The overall R2 values for the ANN, FDA-ANN and SCA-ANN are 0.9951, 0.996 and 0.9955 respectively. The MAE values are 0.099, 0.096 and 0.085 for ANN, FDA-ANN and SCA-ANN respectively. The FDA-ANN model has a higher accuracy than other techniques, such as the ANN and SCA-ANN. The error values obtained for each of the models are very close to their expected value of 0 while their obtained R2 and VAF are also much closer to the targeted value of 1 and 100% respectively. The PI is also close to the expected value of 2. Hence, the three proposed models can be confidently used in predicting RMR values using Q parameters obtained from field investigations without the need to independently determine RMR from the traditional RMR field parameters. The study used the Chord diagram to display the rank of the performance indicators and the sensitivity analysis using the Cosine Amplitude methods (CAM). It shows that the RQD parameter has the highest CAM value followed by Jw and then Jn for all three models. The results offered here provide insight for engineers and academics who are interested in analysing rock mass classification criteria or conducting field investigations.

Sustainable energy consumption in cement production involves practises and strategies aimed at reducing energy use and minimising environmental impact. The efficiency of a cement kiln is dependent on the kiln design, fuel type, and operating temperature. In this study, a dynamic simulation analysis is used to investigate heat losses and distribution within kilns with the aim of improving energy efficiency in cement production. This study used Computational Fluid Dynamics (CFD) with Conjugate Heat Transfer, Turbulent Flow, and the Realisable k−ϵ turbulence model to simulate heat transfer within the refractory and wall systems of the kiln, evaluate the effectiveness of these systems in managing heat losses, and establish the relationship between the heat transfer coefficient (HTC) and the velocities of solid and gas phases. The simulation results indicate that a temperature gradient from the kiln’s interior to its exterior is highly dependent on the effectiveness of refractory lining in absorbing and reducing heat transfer to the outer walls. The results also confirm that different thermal profiles exist for clinker and fuel gases, with clinker temperatures consistently peaking at approximately 1450 °C, an essential condition for optimal cement-phase formation. The results also indicate that phase velocities significantly influence heat absorption and transfer. Lower velocities, such as 0.2 m/s, lead to increased heat absorption, but also elevate heat losses due to prolonged exposure. The relationship between the heat transfer coefficient (HTC) and the velocities of solid and gas phases also indicates that higher velocities improve HTC and enhance overall heat transfer efficiency, reducing energy demand.

The Great Dyke of Zimbabwe is a major geological formation renowned for its rich deposits of platinum group metals. This study addresses the geological and geotechnical challenges faced during mining on the Great Dyke, focusing on the implications for hardrock pillar design. The Great Dyke's geological complexity includes diverse rock types—dunites, harzburgites, pyroxenites, and norites—and notable structural features like joints, faults, and shear zones. These factors complicate the stability of underground workings. Traditional empirical methods and numerical modeling are used in pillar design but fall short in capturing the full complexity of the Great Dyke. The study highlights the absence of advanced methods such as machine learning (ML), artificial intelligence (AI), and geostatistical techniques in current pillar design practices. Incorporating these methods could significantly enhance pillar stability. Geostatistical techniques like kriging offer detailed estimates of rock quality and quantify prediction uncertainty, while ML and AI can analyze extensive data sets to uncover patterns and improve predictions. Integration of real-time data from Industrial Internet of Things sensors into these models allows for dynamic updates and better risk management. Continuous monitoring and adaptive design are essential for maintaining stability in this challenging geological environment. The study's findings aim to guide future mining practices, ensuring enhanced safety and efficiency on the Great Dyke.