Between January and April 2016, three cutting slope failures occurred at Farnley Haugh, Prudhoe and Wylam Scar, along the Newcastle to Carlisle railway line in the Tyne Valley, Northumberland, UK. The valley's superficial geology comprises a complex sequence of Devensian tills, granular glaciofluvial deposits and laminated high-plasticity clays, causing spatial variability in slope stability due to differing soil shear strengths and groundwater conditions. The Farnley Haugh failure was the most significant, severely damaging the railway and causing a line blockage. All three failures occurred during periods of heavy rainfall. Geomorphological and geotechnical assessments showed the original cutting slopes were over-steep with thick granular glaciofluvial deposits exposed on their faces. Forensic slope stability analyses using data from site works, laboratory testing and historical records indicated that before failure, the slopes possessed only marginal stability and low resilience to extreme weather linked to climate change. Existing drainage and vegetation root systems provided insufficient protection. This study demonstrates that substantial sections of Northumberland's aging rail infrastructure remain vulnerable. As climate change intensifies, earthwork deterioration rates will continue to increase. Long-term investment is therefore essential to monitor, maintain and ensure the future safety and resilience of the UK rail infrastructure.
The simplified characterization of rough rock fracture morphology is fundamental to understanding shear mechanical behaviour. This study presents a method for constructing fracture surfaces by integrating profiles with varied inclined dentate asperities. The resulting five surfaces (D_1–D_5) were quantitatively validated using the roughness index θ max ∗ / ( C + 1 ) , demonstrating distinct roughness gradients. Utilizing 3D printing technology, these fracture types were cast into mortar specimens for direct shear testing under varying normal loads. Experimental results allowed a detailed analysis of shear stress, normal displacement, and their correlation with initial roughness and normal stress. Post-test 3D scanning revealed significant spatial variations in macroscopic surface damage. Investigation into the reduction and residual values of surface roughness indicates that degradation is closely linked to initial topography. Specifically, the roughest regions sustain the most significant damage and contribute most substantially to shear resistance. This research offers a feasible and novel approach to evaluating the mechanical properties of rock fractures.
Sonic drilling is a relatively new technique in the UK ground investigation sector. The method has been praised for its continuous core recovery, making it an attractive technology for use in variable stratigraphy, as encountered below London. A comparison of sonic drilled and rotary drilled cores from the Lambeth Group and the underlying Chalk Group was conducted using data from HS2 on the ground investigation stages of the Northolt Tunnel project. A comparison has been made between core recovery, sample quality and laboratory testing data within the encountered strata. Sonic drilling performed well in providing near-complete core recovery throughout the Sand Channels, Mottled Clay and the Chalk Group relative to the rotary boreholes. Differences were observed in the recovered fabrics of the Lambeth Group, with rotary drilling preserving more of the finer geological structures. In the Chalk Group, the sonic drilling technique remoulded the chalk to a putty, and therefore solid core recovery and sample quality were far higher in the rotary-drilled boreholes. This comparison highlights that the strengths and limitations of drilling technologies must be understood to maximise their potential and drilling methods should be tailored to the ground conditions and the geotechnical objectives to optimise the data received.
Surface horizontal well staged hydraulic fracturing is an effective technique for preventing and controlling rockbursts induced by hard roofs in coal mines. For working faces with excavated roadways, the induced fracture network not only fully cover the hard roof across the entire working face but also avoid uncontrollable secondary damage to the roadways. Using a working face in Zhaoxian Coal Mine, Shaanxi Province, China, as a case study, this study investigates the key technologies for rockburst-prevention horizontal well fracturing from three aspects: optimization of the target key stratum within the thick hard roof, control of the horizontal well trajectory, and optimization of fracturing parameters. By integrating structural analysis of overburden strata, identification of rockburst-inducing layers, evaluation of energy release characteristics, and measurement of the development heights of the caving zone and water-conducting fracture zone, the target key stratum was determined to be located 51–64 m above the coal seam roof. Accordingly, the horizontal well was positioned within the siltstone layer at approximately 51 m above the coal seam roof. Based on geological features revealed by the roadways, the dip angle of the working face was corrected in real time, and a geological-guided model was developed to precisely control the horizontal section trajectory within the siltstone layer. Furthermore, parameter optimization ensured that the fracture network effectively covered the entire hard roof area while minimizing secondary roadway damage. Microseismic monitoring confirmed fracture heights of 52–55 m and lengths of 304–335 m, successfully covering the target roof. During coal extraction, 89% of monitored microseismic events exhibited low energy levels (0–10 3 J), with no high-energy events (≥10 4 J), demonstrating the effectiveness of the proposed rockburst prevention method. These results provide both theoretical support and engineering guidance for the optimal design of staged hydraulic fracturing in rockburst-prevention horizontal wells under excavated roadway conditions.
The rapid urban development in eastern Tabriz, Iran, has highlighted the geotechnical challenges posed by widespread argillaceous marly soft rocks that exhibit complex mineralogy and weak mechanical behaviour. These Neogene sediments, deposited in a shallow, low-energy lacustrine environment, contain gypsum, illite, muscovite and fossil fragments. Their high porosity, low cementation and heterogeneous microstructure contribute to their limited strength and durability, characteristics that are inadequately captured by conventional rock mechanics tests. This study employs an integrated approach combining soil and rock mechanics methods, including X-ray diffraction (XRD), scanning electron microscopy (SEM), petrographical analysis and standard mechanical testing, to systematically characterize the mineralogical, microstructural and geomechanical properties of these transitional deposits. Results demonstrate that clay bonding, carbonate cementation and microstructural heterogeneity critically control strength, stiffness and failure behaviour. To improve the engineering assessment and material classification, two indices – the strength index (STI) and the destruction index (DEI) – are introduced. Materials with STI < 0.5 MPa and DEI > 30% are highly prone to instability, whereas those with STI > 1 MPa and DEI < 10% display a favourable geomechanical performance. The proposed classification framework provides a robust tool for evaluating weak argillaceous rocks, supporting more reliable and cost-effective foundation design in similar geological settings.
Cornwall's history as one of the richest metal-mining regions in the world has left a legacy of hundreds of kilometres of abandoned mine workings containing millions of cubic metres of water. Despite the growing development of mine-water heat schemes in former coal-mining areas in the UK, the potential geothermal resource within the flooded mines in Cornwall has yet to be fully recognized. The utilization of this resource presents a few challenges that are different from those met when extracting heat from former coal mines, but it presents readily accessible opportunities for producing sustainable low-carbon heating, cooling and thermal storage. There are recorded to be 154 mines in Cornwall that are over 200 m deep. Recent studies have identified the temperature and volumetric resource that may be available in some of the deep mines, together with examples of four mines that show distinctly different types of hydrogeological systems. This has demonstrated the potential and the importance of thoroughly understanding the hydrodynamics of the shafts and underground workings, and how these can be best utilized as a low-enthalpy heat resource.
As mudrocks underlie large areas of the UK, including major urban areas and infrastructure routes, they frequently form the bedrock at construction sites and slope and tunnel engineering works. They may also be used as fills or construction materials. They range widely in terms of strength and compressibility, depending on composition, degree of compaction and induration, weathering grade, fabric and structure. A review of the performance of mudrocks in engineering situations including slopes, tunnels and in embankment and foundation construction highlights the importance of durability and the role of chemical weathering processes, especially the oxidation of pyrite, in the engineering performance of these materials. It is necessary for the effects of changes in the environment of mudrocks, brought about both during and after completion of engineering works, to be anticipated early in projects so that design and construction measures that circumvent or avoid problems can be adopted. Advances in the understanding of factors controlling mudrock durability that can facilitate improved prediction of changes in properties are described. It is argued that understanding controls on the properties of mudrocks underlies ways of avoiding the problems that can arise in civil engineering works involving these materials. It is apparent that there is much to be gained from the publication of case histories of past engineering projects, particularly those in which problems stemming from unexpected changes in the properties of mudrocks have occurred.
Across Great Britain an estimated 25% of all addresses are located above a coalfield. Decarbonizing heating for buildings is a key challenge, with one technology-ready solution being heat networks utilizing mine water from flooded underground coal mines. However, there have been challenges at a strategic planning level identifying areas with the best potential for mine water heat. To support local area energy plans and pilot heat network zoning projects, we describe a method to map mine water heat opportunities at a scale suitable for high-level assessments. Datasets including depth to workings, mine water levels, geological disturbances, mine water treatment schemes and discharges were used, with areas ranked into 'Good', 'Possible' and 'Challenging' opportunities for open-loop mine water heat schemes. The method is illustrated with a case study from part of a national opportunity assessment completed across the coalfields of Wales. The case study area covers Caerphilly County Borough Council, where 24.6% of addresses were located in 'Good' opportunity areas. Opportunity maps were also created for the following cities in England: Birmingham, Bristol, Coventry, Leeds, Manchester, Newcastle, Nottingham, Sheffield, Stoke-on-Trent and Sunderland. Data limitations include a lack of fully digitized mine plans and areas with limited hydrogeological information, and the approach itself could be easier to repeat if more automation was included. The opportunity maps have proved useful for stakeholder engagement, raising the profile of mine water heat and integration into heat network planning; however, they do not remove the need for site-specific feasibility studies.
Under long-term multiaxial stresses, rock masses exhibit gradient stress characteristics. Investigating fracture precursors and crack propagation under such conditions is crucial for monitoring and preventing rock engineering disasters. This study analysed rock behaviour using acoustic emission (AE) and digital image correlation (DIC) techniques under a gradient stress field. The findings indicate that as the gradient stress difference grows, rock strain decreases and the effective elastic modulus decreases. The load–displacement curve exhibits a dual-peak fluctuation pattern, with an increasing interval between the peaks and opposite evolution trends in the high-stress and low-stress zones. A larger gradient stress difference extends the ‘warning window’ before failure, enhances macroscopic crack precursors and increases the total crack count, with tensile cracks decreasing and shear cracks increasing linearly but at a slowing rate. Fracture patterns shift from block-like to fragmented. Higher gradient stress differences increase the proportion of total energy in the high-stress zones, with both zones experiencing more dissipated energy and less elastic energy, although dissipation remains higher in the low-stress regions. This work advances the understanding of rock deformation, failure modes and fracture characteristics in gradient stress settings, supporting future research.
Debris flow susceptibility depends on complex factor interactions; traditional single models have limited accuracy and generalization in complex geological settings. This study develops a stacking heterogeneous ensemble learning model based on watershed units, integrating 10 evaluation factors. It systematically compares five ensemble configurations and a single XGBoost model, analysing driving mechanisms via Shapley additive explanations (SHAP) values. Results show that the optimal stacking ensemble model, which uses linear SVM, XGBoost and Gaussian naive Bayes as base models and logistic regression as the meta-model, achieves an the area under the receiver operating characteristic curve (AUC) of 0.927. Zones with extremely high susceptibility cover 24.24% of the area. SHAP values reveal that intensive human engineering activities within 5 km of roads significantly enhance susceptibility by degrading vegetation and soil structure. Annual precipitation is the key water supply factor, whereas high-altitude snowmelt inhibits susceptibility. The topographic wetness index in the range of 6.5-10, Melton ratio in the range of 700-1000 and river network density in the range of 0.5-1.5 synergistically regulate susceptibility via the coupled processes of water convergence, flow dynamics and channel evolution. This study integrates the optimized stacking model's high-precision prediction and mechanism interpretation, establishing a novel framework for debris flow susceptibility assessment in complex geological settings.
Despite having limited cave development, the UK Permian, Jurassic and Cretaceous carbonates have standard karst features - caves, stream sinks, dolines, springs, dry valleys, and solutional networks of fissures and conduits extending multiple kilometres. This paper provides a new discussion and comparison of these karst aquifers, demonstrating that significant karst features are widespread. Tracer tests in the Chalk and Jurassic limestones demonstrate flows of 1000s, m/day from the surface to a spring or abstraction over many kilometres, comparable to results from karst aquifers with substantial cave systems, such as the UK Carboniferous limestones. They are less vulnerable in terms of having less cave development and a lower proportion of rapid point recharge, with greater attenuation via flow through smaller voids. However, the combination of long-distance transport via karstic networks and long-term storage of contaminants in smaller voids may make contaminant management particularly challenging. Karst development and the proportion of rapid recharge vary substantially at different spatial scales. Therefore, better groundwater conceptual models incorporating karst data at the local catchment scale, together with further investigations (tracer tests, borehole studies, water balance) are needed. Groundwater management and protection require the use of methods appropriate to the karstic nature of these aquifers.
The structure of loess is a key factor influencing its mechanical and hydraulic properties. Research on structural evaluation indices has been increasing, and existing studies, each with their own distinct contributions, have greatly advanced the understanding of soil structure. However, most approaches rely on testing undisturbed, remoulded and saturated samples, which involve complex procedures and numerous variables. Moreover, it remains difficult to accurately quantify the individual contributions of different influencing factors to soil structure. Therefore, developing a well-defined, easily measurable and parameter-efficient structural index for loess that also accounts for hydro-mechanical coupling represents a significant goal in current research. In this study, a transient vacuum decay air permeameter was used to investigate the air permeability of loess under varying moisture conditions and hydro-mechanical coupling, and to explore its relationship with structural characteristics. The results show that for undisturbed loess of different textures, air permeability increased as water content decreased during drying, and decreased as water content increased during wetting; with drying permeability consistently higher than wetting permeability. At the same water content, air permeability decreased with increasing vertical pressure. For the same soil texture, a higher water content led to lower air permeability under the same pressure and air-filled porosity. Based on these findings, a novel expression has been proposed to describe the variation of the structure ratio with water content and pressure. Furthermore, using the Poulsen model, predictive analyses were conducted to relate structural parameters to moisture and air-filled porosity, as well as to moisture, pressure and air-filled porosity. This provides a new approach for the quantitative characterization of loess structure. As the determination of air permeability has obvious advantages, such as being non-destructive, efficient, economical, convenient and fast, it has the unique advantage of being used as an index for evaluating the structural properties of loess and is worth being vigorously popularized.
Understanding historical landslide occurrence, distribution and impacts can help to inform modern risk-mitigation strategies. By reviewing the National Library of Wales' recently digitized catalogue, a database of 532 unique landslide events was compiled spanning almost 100 years from 1822 to 1919. The mass movements were catalogued and geolocated with extent and timing information, as well as failure mode, impacts, antecedent weather and trigger mechanisms. A wide range of transport infrastructure, commercial activities and residential buildings were impacted. A total of 119 fatalities were identified. The distribution of events is compared with that of existing landslide databases for Wales. Landslides were found to occur predominantly in winter, and the dataset shows an apparent increase in landslide frequency from 1870 to 1909. The paper explores the links between landsliding and industrial development, as well as weather events. Early approaches to landslide mitigation and remediation are considered. Select case study sites are examined in detail where landsliding remains an on-going hazard.
Mine water geothermal schemes (MWGS) typically involve two or more boreholes and require power to pump water from depth, incurring exploration, drilling and operational costs. Gravity-driven mine water surface discharges are a common feature of former coal-mining areas. Where discharge flow rates are high enough, MWGS could be installed to provide low-cost, low-carbon heating. Situated in an urban environment, the Old Fordell mine water discharge is one of the largest in Scotland. With a total iron load of 130 t a -1 , it also represents a significant environmental concern for the receiving watercourse. To determine potential heat output, a synthetic analysis was performed using legacy mine records and hydrogeological monitoring (chemistry, temperature, flow, water level measurements) data from the UK Mining Remediation Authority. Between 2007 and 2019, regional mine water levels rose from −120 m ordnance datum (OD) to +40 m OD before stabilizing. During this process, Old Fordell temperature rose by 2°C and chemical composition changed, indicating introduction of deeper-sourced water. A heat extraction Δ T of 8°C, mean discharge temperature (12.6°C) and flow rate (98 l s -1 ) yields 4.4 megawatts (MW), enough to heat c. 1100 homes. The potential heat output is deemed sustainable as long as regional mine water levels remain stable, showing that such features can supply heat to local users and have dual benefit of offsetting costs associated with installation and operation of any future mine water treatment plants.
Two internationally approved methods for attaining a ground model now exist, that of the International Association of Engineering Geology (Commission 25) and that of Eurocode 7 (2nd Generation). The similarities and differences between these two are sufficient to cause confusion in practice and the purpose of this note is to highlight their key differences and provide guidance on the safe use of both systems. Shared vocabulary is clarified and the basic philosophy of both systems explained and compared. The problems that can arise when these differences are not appreciated are explained and recommendations to help avoid these problems provided for those who have to procure a ground model, for those who have to generate such a model and for those who have to use them in design and construction, and possibly during the working life of the engineering to which they are related.
Mining-induced landslides in southwestern China are being more frequently reported, but quantifying the erosion-entrainment effect on their mobility requires further attention. This study reconstructs the 3 December 2004 Zuojiaying landslide in Guizhou, China, where a rainfall- and mining-triggered rock collapse evolved into a debris slide that claimed 44 lives. A field investigation and an uncrewed aerial vehicle (UAV)-based digital elevation model are combined with Rapid Mass Movement System (RAMMS) simulations to reproduce the landslide's dynamic characteristics and assess the effect of entrainment on runout. Back-calculated models suggest that the landslide volume increased from an initial 11 000 m(3) to c. 36 000 m(3) due to entrainment. A sensitivity study shows that denser and more saturated erodible material in the transport zone substantially increases the entrained volume. A comparison of scenarios with and without entrainment reveals that runout extended from c. 380 to c. 430 m, with the maximum deposit thickness increasing from c. 4 to c. 7 m. A stage-wise analysis of the runout process emphasizes the critical role of erosion-entrainment in regulating mobility and the deposit characteristics of landslides, which includes four stages from rock collapse to debris slide. These findings provide quantitative constraints on substrate erodibility and entrainment that govern the runout and deposit thickness, improving hazard assessment and zoning in areas with similar conditions.
In deep metal mines, repeated blasting during ore extraction causes progressive damage to surrounding rock masses that accumulates under long-term loading and may ultimately lead to rock mass instability. This study investigates the damage characteristics of the overlying rock mass under static and dynamic loads through integrated laboratory testing, physical modelling and FLAC3D numerical simulations. Taking the large-scale mining operations at Luohe Iron Mine as a case study, the dynamic damage evolution of the overlying rock mass is examined under multi-panel, multi-stope mining conditions. Results indicate that one-step mining produces minimal damage and subsidence in the surrounding rock mass, whereas two-step mining significantly increases both the extent of damage and the magnitude of subsidence. The most severe damage occurs at the inter-layer rock beams and at the roof and floor of mining panels. This intensified damage reflects the combined effects of frequent dynamic loading and inter-layer damage superimposition, which together exceed the damage produced under static loading alone and become the dominant factor controlling rock mass damage. Under sustained loading, blast-damaged rock masses exhibit a characteristic deformation pattern, converging inward from the lateral boundaries toward the central floor.
Real-time prediction of rock slope stability in active mines remains a critical challenge due to complex geology, dynamic mining stress and environmental factors. The Pulang Copper Mine, with its complex structural setting and ongoing subsidence, requires advanced monitoring to mitigate failure risk. The aim of this study was to develop and validate an Internet of Things (IoT)-driven Enhanced Transformer model for real-time prediction of the factor of safety (FoS) and stability classification, integrating numerical simulation, IoT data streaming and deep learning to improve the early-warning capability. A FLAC3D simulation replicated 2 years of mining (730 daily steps) at six strategic monitoring points, generating time-series data for displacement, velocity, acceleration and FoS. An IoT framework streamed these data with less than 5 s latency. An Enhanced Transformer architecture with multi-head self-attention, multi-task learning ( lambda = 0.5) and advanced feature engineering was trained on the sequences. The Enhanced Transformer achieved a superior performance, with a testing coefficient of determination (R-2) ranging from 0.416 (Station 5, characterized by complex transitional kinematics) to 0.991 (Station 4). The testing mean absolute error (MAE) ranged from 0.003596 (Station 2) to 0.019071 (Station 5), a reduction of up to 88% compared to the Standard Transformer. For four-class stability classification, the model attained a mean test accuracy of 0.993, with the critical-class recall reaching 1.0 - guaranteeing zero missed alarms for life-threatening critical and unstable conditions, the paramount objective for early-warning systems. The proposed IoT-Enhanced Transformer model provides a highly accurate, real-time solution for slope stability prediction, significantly outperforming conventional models.
