RSS 1.0Mining and Geological Engineering
Collections in this community
Recent Submissions
-
Open-pit slope geotechnical considerations and its effects on mine planningIn open-pit mining, a stable pit slopes design is essential for safe operation and economic performance of the mine. However, a steeper pit is more desirable from an economic standpoint due to reduced overburden removal. As the mine deepens, the open-pit walls become increasingly prone to slope failure, which causes human and economic losses. Therefore, a feasible and stable slope mine design requires a serious geotechnical investigation. The optimization of this design requires steepening the overall slope angle as much as possible while maintaining mine safety for efficient and effective mining operations. The open-pit slope geotechnical investigation calls for detailed geological and geotechnical data and advanced numerical modeling. In this study, geological and geotechnical data are collected from the Erdenet Copper Mine of Mongolia. The collected information includes data from discontinuity face mapping, geotechnical core logging, groundwater condition, geological exploration cross-sections, pit map, and rock property lab test results. The open-pit slope stability is analyzed with geotechnical numerical modeling software FLAC2D, and the variation and distribution of factors of safety (FOS) are computed and studied. The stability of Erdenet mine’s North-West open-pit is simulated by dividing the pit into ten representative cross-sections, and subsequently, FOS is calculated for each cross-section. The simulation results show that each cross-section has a higher overall FOS value than the allowable mine FOS, set at 1.5 with an earthquake magnitude of 0.165g peak ground acceleration (PGA). However, the localized high shear strain on individual benches may still occur, which can cause potential failures. Parametric studies indicate that changes in the bench angles and rock mass properties will have various degrees of impact on pit slope FOS. The effect of bench angle changes appears to be more significant. The study of pit slope design on mine planning shows that a 1° increase on slope angle will reduce excavation volume by 5 M m3 and save $15 million in excavation cost, but will also reduce FOS by 0.12. Engineering judgment and decision will have to be made regarding this tradeoff for a safe and economical mining operation. Practice and analysis indicate that the computer simulation alone is not sufficient to ensure the accurate estimation of slope stability. It is recommended to use a combination of slope monitoring and computer simulation to provide verification against each other to detect any potential hazards in mine. Mine pit slope movement monitoring program setup and monitoring procedure are analyzed and proposed in this study. The above findings allow mining engineers to optimally design pit slopes under the given geotechnical conditions and minimize the risk of slope failures while improving the stripping ratio and enhancing production profit.
-
Engineering economic analysis of a rail extension from Dunbar siding to Livengood, AlaskaThe Dunbar Siding to Livengood rail extension study is an economic prefeasibility investigation, and is conducted from two perspectives as a cost benefit analysis. The first perspective is, that of the Alaska Railroad Corporation (ARRC) in which the capital and operating costs of the proposed extension are recovered through the revenue stream resulting from the out-bound mineral freight loads, the in-bound re-supply freight loads, and the potential commuter passenger service to mining projects and communities in the Livengood area. The second perspective is that of the private sector in which a shipping sensitivity and employee transport analysis with respect to mining project developments. The large mineral resource base within the Dunbar-Livengood Corridor indicates an excellent freight potential with generous benefits for Alaska's economy of greater than $2 billion annually in gross revenues; whereas, resource and rail development are synergistic.
-
Laboratory investigation of infiltration process of nonnewtonian fluids through porous media in a non-isothermal flow regime for effective remediation of adsorbed contaminantsContamination of soil and groundwater have serious health implications for man and environment. The overall goal of this research is to study a methodology of using nonNewtonian fluids for effective remediation of adsorbed contaminants in porous media under nonisothermal flow regimes. Non-Newtonian fluids (Guar gum and Xanthan gum solutions) provide a high viscous solution at low concentration and these fluids adjust their viscosities with applied shear rate and change in temperature. Adjustment of viscosity with an applied rate of shear is vital for contaminant remediation because non-Newtonian shear thinning fluids can penetrate to low permeability zones in subsurface by decreasing their viscosities due to high shear rates offered by low permeability zones. The application of non-Newtonian shear thinning fluids for contaminant remediation required the improvement in understanding of rheology and how the factors such as concentration, temperature and change in shear rate impacted the rheology of fluids. In order to study the rheology, we studied the changes in rheological characteristics (viscosity and contact angle) of non-Newtonian fluids of different concentrations (i.e., 0.5g/l, 1g/l, 3g/l, 6g/l and 7g/l) at different temperatures ranging from 0 ºC to 30 ºC. OFITE model 900 viscometer and Tantec contact angle meter were used to record the changes in viscosity of fluids for an applied range of shear rate (i.e., 17.02 s⁻¹ to 1021.38 s⁻¹) and contact angles, respectively, for different concentrations of non-Newtonian fluids. Understanding the flow characteristic of non-Newtonian fluids under low temperature conditions could help in developing methods to effectively remediate contaminants from soils. Results of rheological tests manifested an increase in the viscosity of both polymers with concentration and decrease in temperature. Mid (i.e., 3g/l) to high (i.e., 6g/l and 7g/l) concentrations of polymers manifested higher viscosities compared to 0.5g/l for both polymers. Flow of high viscous solutions required more force to pass through a glass-tube-bundle setup which represented a synthetic porous media to study the flow characteristic and effectiveness of Newtonian and non-Newtonian fluids for contaminant remediation. Low concentrations of 0.5g/l were selected for flow and remediation experiments because this concentration can flow through porous media easily without application of force. The 0.5g/l of Xanthan gum and de-ionized water were used to conduct the infiltration experiments to study the flow characteristics of Newtonian and non-Newtonian fluids at 0.6°C, 5°C and 19°C in synthetic porous media. Infiltration depth of both Newtonian and non-Newtonian fluids would decrease with the decrease in temperature because of the change in their properties like dynamic viscosity, density and angle of contact. The result of comparison of Newtonian and non-Newtonian fluids showed water to be more effective in remediating a surrogate adsorbent contaminant (Dichlobenil) from the synthetic porous media at 19°C. This result was counter-intuitive to what we began with as our hypothesis. However, it was also observed later that 0.5 g/l concentration of Guar gum behaved more like a Newtonian fluid and 0.5 g/l concentration of Xanthan gum had not shown strong non-Newtonian behavior compared to higher concentrations of Xanthan gum. Hence more analysis needs to be done to determine what concentration of non-Newtonian fluid should be more effective for remediation.
-
Enhancement of algorithm for detection of gold strip circuit vessel sensor errorsSensors are used to understand the condition and flow of mineral processes. Having accurate and precise information is fundamental for proper operation. Even small errors are relevant to cost when considering the operational span of a mine. Finding small errors is hard; few algorithms can detect them and fewer still, when considering errors on the scale of 2% in magnitude. Some tools have recently been developed using data mining techniques for detecting small errors. Rambabu Pothina (2017) created an algorithm for detecting small errors in strip vessel temperature sensors in the carbon stripping circuit in Pogo mine. The algorithm performed well and was able to detect small magnitude errors without disrupting the industrial process. This thesis improves the understanding of the performance of the algorithm, while also making some minor changes. First, a statistical analysis of the results of the algorithm on baseline data revealed an inherent difference in how the carbon strip process was run with respect to the two strip vessels. This discovery provided insight into the algorithm, and how its performance depended on process characteristics. Second, the error detection algorithm was tested under scenarios different from Pothina (2017). Three separate types of errors were artificially added to real data: a) a fixed 2% ("fixed" error increase) b) a fixed 2% decrease ("fixed" error decrease) and c) an error with a mean value of 2% of magnitude ("noisy" error). Additionally, error was added to temperature data from each strip vessel (rather than just one), though only one at a time. The algorithm was tested under each scenario for each of the four years, 2015, 2016, 2017 and 2018. The time to detect errors ranged from 19 to 73 days. The time to detect was very high (53 to 73 days) in 2017 since there were large data gaps that year. In general, time to detect was about 30 days. The performance under noisy error were not that far below fixed error scenario. The algorithm took 10% more time to detect errors under noisy error scenario compared to fixed error scenario. On average, the algorithm detected an error after 25 cycles, regardless of the time span this represents. This is consistent in years with plentiful data, such as 2015, as well as years with less data, 2017 and 2018. In years with data gaps, 25 cycles represent a longer time period. Seeded errors that decreased vessel temperature have very similar results to its equivalent increase, i.e. the decrease in 2% of S2 has results similar to the increase of 2% in S1 and vice versa. In conclusion, these additional testing and analysis helped develop a more comprehensive understanding of the behavior of the data and the algorithms. These results validate and strengthen the findings of Pothina (2017).
-
Enhanced bioweathering of coal for rare earth element extraction and concentrationRare earth elements (REEs) are a group of seventeen elements that include scandium, yttrium, and fifteen of the lanthanide series elements, which are used in a variety of consumer goods and for defense purposes. Acquiring a domestic profitable source of REEs is a critical national need as most of the global supply comes from one country, China. To counter this problem, the US is actively looking at alternative sources of REEs by implementing unconventional methods of extraction. Coal is one of the alternative sources of REEs. Alaskan coal from Wishbone Hill and Healy are known to contain REEs up to 286 ppm and 524 ppm, respectively, while having concentrations as high as 950 ppm on ash basis in some density fractions. Microbial leaching or bioleaching is a novel method that can be used for extraction of REEs from coal as microbes are known to affect earth's surface over geologic time by playing critical roles in weathering of minerals. A certain species of bacteria, Shewanella oneidensis MR-1, was used to separate the REEs from Wishbone Hill and Healy coal samples. The experiments were performed for various density fractions of both coals by varying solids percentage, temperature, size of coal, and bacterial concentration, and recovery of REEs for these conditions was recorded. Highest individual recovery of neodymium, 75.3%, was obtained for Wishbone Hill 1.3 floats, while a maximum of 98.4% total REE recovery was obtained for Healy 1.3 sinks. Healy coal has the higher total recovery of REEs in comparison to Wishbone Hill coal. Bioleaching process was also compared to the acid leaching process. Healy coal responded better to bioleaching than the acid leaching process. The Wishbone Hill coal had comparable recoveries of bioleaching with acid leaching, although they were always less than acid leaching.
-
Analysis of cation-treated clay microstructure using zeta potential and x-ray diffractionUnfrozen water within cation-treated, fine-grained frozen soils is a key element in cold regions engineering, and is heavily influenced by the surface charge of the soil's clay component. This study investigated the effects of the surface charge of cation-treated clay soils by measuring the zeta potential as a function of temperature, and measuring changes in the micro-structure of frozen cation-treated clays using the x-ray diffraction (XRD) method. I tested five treatments (untreated, and Ca²⁺, Mg²⁺, Na⁺, and K⁺ treatments) of six soils (montmorillonite, kaolinite, illite, illite-smectite, chlorite, and Copper River soil). The zeta potential demonstrated a negative relationship with temperature change for both above-freezing and sub-freezing conditions (-1 to 20 °C). Temperature had a greater effect on the monovalent-treated soils that contain smectite minerals, which included montmorillonite, illite-smectite, and the Copper River soil. Monovalent cation-treated soils demonstrated large negative trends and more negative zeta potential, whereas divalent cation-treated soils demonstrated less negative trends that were less dependent on temperature. The cation treatment will affect the Debye-length, also affecting the zeta potential and arrangement of clay particles. More negative zeta potential (i.e., soil dominated by monovalent cations) will lead to a dispersed structure, whereas less negative zeta potential (i.e., soil dominated by divalent cations) will lead to a flocculated structure. XRD research indicated that the montmorillonite samples demonstrated decreased dspacing compared with the International Center for Diffraction Data (ICDD) standard. The K⁺- treated montmorillonite, untreated montmorillonite, and untreated illite-smectite samples demonstrated donut-shaped pole figure results, which may indicate that the results are an artifact of sample preparation rather than a reflection of the cation effects on the structure of the clay. Improved could be made in sample preparation to eliminate ice lens formation during freezing, which may improve the success with the XRD method. Scanning electron microscopy (SEM) should be used to observe the frozen clays, especially montmorillonite, illite-smectite, and the Copper River soil, as it may reveal the internal geometry of voids and the possible relationship between ice and the clay structure, increasing our understanding of the clay structure at the microaggregate scale.
-
The mechanics of diamond core drilling of rocksIn an attempt to study the mechanics of diamond core drilling in rocks, an investigation on rock drillability was conducted at the University of Alaska Fairbanks. A series of drilling and coring tests was conducted on six types of rock using several different diamond bits. Factors involved in a diamond coring and drilling process such as weight-on-bit, rotational speed, and rock type were identified and the effects of those parameters were experimentally evaluated based on the penetration rate, applied torque, and specific energy. Statistical techniques were used to design the drilling tests and to develop drilling models. Fundamentals of rock failure mechanics in relation to rock drilling were reviewed. Several existing rock drilling models were also examined with the data from this study. Results indicated that all of the three drilling parameters, i.e., the penetration rate, applied torque, and specific energy, were significantly affected by the weight-on-bit and rock type. The penetration rate of a bit was also affected by the rotational speed. The effects of the rotational speed on the applied torque and specific energy, however, were found to be insignificant. It was also found that the theoretical models can be used to predict the maximum effective weight-on-bit and penetration rate. Among the four theoretical models examined, the elastic model predicted the most accurate penetration rate. The maximum effective weights-on-bit predicted by the plastic model and the two fracture models, however, were close to each other and in agreement with the experimental observation. Statistical models developed in this study were used to predict the penetration rate in the Rock Drilling under the Greenland Ice Sheet project. The variation between the predicted value and the actual value was less than 10%.
-
Comprehensive Investigation Into Historical Pipeline Construction Costs And Engineering Economic Analysis Of Alaska In-State Gas PipelineThis study analyzes historical cost data of 412 pipelines and 220 compressor stations. On the basis of this analysis, the study also evaluates the feasibility of an Alaska in-state gas pipeline using Monte Carlo simulation techniques. Analysis of pipeline construction costs shows that component costs, shares of cost components, and learning rates for material and labor costs vary by diameter, length, volume, year, and location. Overall average learning rates for pipeline material and labor costs are 6.1% and 12.4%, respectively. Overall average cost shares for pipeline material, labor, miscellaneous, and right of way (ROW) are 31%, 40%, 23%, and 7%, respectively. Regression models are developed to estimate pipeline component costs for different lengths, cross-sectional areas, and locations. An analysis of inaccuracy in pipeline cost estimation demonstrates that the cost estimation of pipeline cost components is biased except for in the case of total costs. Overall overrun rates for pipeline material, labor, miscellaneous, ROW, and total costs are 4.9%, 22.4%, -0.9%, 9.1%, and 6.5%, respectively, and project size, capacity, diameter, location, and year of completion have different degrees of impacts on cost overruns of pipeline cost components. Analysis of compressor station costs shows that component costs, shares of cost components, and learning rates for material and labor costs vary in terms of capacity, year, and location. Average learning rates for compressor station material and labor costs are 12.1% and 7.48%, respectively. Overall average cost shares of material, labor, miscellaneous, and ROW are 50.6%, 27.2%, 21.5%, and 0.8%, respectively. Regression models are developed to estimate compressor station component costs in different capacities and locations. An investigation into inaccuracies in compressor station cost estimation demonstrates that the cost estimation for compressor stations is biased except for in the case of material costs. Overall average overrun rates for compressor station material, labor, miscellaneous, land, and total costs are 3%, 60%, 2%, -14%, and 11%, respectively, and cost overruns for cost components are influenced by location and year of completion to different degrees. Monte Carlo models are developed and simulated to evaluate the feasibility of an Alaska in-state gas pipeline by assigning triangular distribution of the values of economic parameters. Simulated results show that the construction of an Alaska in-state natural gas pipeline is feasible at three scenarios: 500 million cubic feet per day (mmcfd), 750 mmcfd, and 1000 mmcfd.
-
Multiphysics Modeling Of Gaseous Contaminant Transport In Deep Open Pit Mines Under Arctic Air InversionsEntrapment of pollutants in a deep open pit operating in a cold climate could occur due to atmospheric inversion. The process of air inversion is complex and requires thorough understanding in order to design a mine ventilation plan to remove trapped pollutants in open-pit mines operating in the arctic/sub-arctic regions. The objective of this dissertation is to develop a model using Computational Fluid Dynamics (CFD) tools for analysis of gaseous pollutant transport in deep, open pit mines under air inversion in arctic or subarctic regions. An Eulerian 3-D model was used for the development and validation of the CFD model of pollutant transport in an idealized open pit mine. No prior assumptions, turbulent or laminar, were considered for the nature of the flow. The 2-D model results indicated that air velocity, air temperature, diffusivity coefficient and slope angle were important controlling parameters in the inversion process. The flow regime was laminar at the origin, but as the flow progressed toward the center of the pit it changed to quasi-laminar and generated local eddies towards the pit bottom. The total energy of the quasi-laminar flow as well as the small local eddies was not enough to lift the inversion cap. However, a combination of quasi-turbulent flow and the local eddy transport resulted in removal of some of the pollutant mass from the pit bottom, either due to turbulent mixing, or due to advection. Presence of backflow may appear to be a logical mode of flow in deep open-pit mines in arctic regions. Next, the 3D model was validated using data from a selected open-pit mine. Influent air velocity, diffusivity coefficient, larger pit geometry were found to influence the retention and transport of pollutant out of the pit. The most important conclusion that was drawn from this research is that natural ventilation alone cannot remove the pollutants from an open pit or lift the inversion cap.
-
Experimental And Numerical Study Of Sonic Wave Propagation In Freezing Sand And SiltA numerical model for delineating the temperature-velocity relationship of freezing porous media and soil is developed in Matlab based on Leclaire's Biot-type three-phase theory. Leclaire's theory gives lower sonic velocities than the experimental results because it does not take into consideration the effect of the solid-ice frame when water is freezing. To take the solid-ice effective frame into account, the average bulk and shear moduli estimation are modified with a proposed procedure. The modification gives higher P-wave and S-wave velocities that fit experimental data well. A comprehensive suite of physical and acoustic laboratory experiments are conducted on artificial sands, sand-clay mixtures and Fairbanks silts to investigate the temperature-velocity relationship during the freezing process and the effects of grain size and fine clay content. A Multi-channel ultrasonic scanning system (MUSS) is designed, installed and programmed for the experimental computerized ultrasonic tomography (CUST) study. The inward and outward freezing process and freezing front development in Fairbanks silt samples are observed using computerized ultrasonic tomography (CUST) in the laboratory. The experiments generate sonic wave velocity and temperature distribution during the freezing process. The freezing front is clearly identified in the CUST as a function of time and temperature. Comprehensive numerical finite element method (FEM) simulations, which account for the conduction in porous media, the latent heat effect and the nonlinear thermal properties of soil, are performed on the inward and outward freezing process of Fairbanks silt based on the experimental conditions. In conjunction with the temperature-velocity model developed in the study, sonic wave velocity tomograms are generated. The results are comparable with those obtained by CUST. The study indicates that CUST is an effective method for studying freezing processes and has potential for indirect measurement of unfrozen water content variations in the soil without interfering with the freezing process.
-
Modeling Of Depressurization And Thermal Reservoir Simulation To Predict Gas Production From Methane -Hydrate FormationsGas hydrates represent a huge potential future resource of natural gas. However, significant technical issues need to be resolved before this enormous resource can be considered to be an economically producible reserve. Developments in numerical reservoir simulations give useful information in predicting the technical and economic analysis of the hydrate-dissociation process. For this reason, a commercial reservoir simulator, CMG (Computer Modeling Group) STARS (Steam, Thermal, and Advanced Processes Reservoir Simulator) has been adapted in this study to model gas hydrate dissociation caused by several production mechanisms (depressurization, hot water injection and steam injection). Even though CMG is a commercially available simulator capable of handling thermal oil recovery processes, the novel approach of this work is the way by which the simulator was modified by formulating a kinetic and thermodynamic model to describe the hydrate decomposition. The simulator can calculate gas and water production rates from a well, and the profiles of pressure, temperature and saturation distributions in the formation for various operating conditions. Results indicate that a significant amount of gas can be produced from a hypothetical hydrate formation overlying a free gas accumulation by several different production scenarios. However, steam injection remarkably improves gas production over depressurization and hot water injection. A revised axisymmetric model for simulating gas production from hydrate decomposition in porous media by a depressurization method is also presented. Self-similar solutions are obtained for constant well pressure and fixed natural gas output. A comparison of these two boundary conditions at the well showed that a higher gas flow rate can be achieved in the long run in the case of constant well pressure over that of fixed gas output in spite of slower movement of the dissociation front. For different reservoir temperatures and various well boundary conditions, distributions of temperature and pressure profiles, as well as the gas flow rate in the hydrate zone and the gas zone, are evaluated.
-
An Evalulation Of Variables Affecting Gold Extraction At A Mineral Processing Plant Operated In A Sub-Arctic EnvironmentThe Fort Knox Mine, located 25 miles northeast of Fairbanks, Alaska, is operated in a sub-Antic environment. Since process slurry temperatures cycle seasonally with air temperature, the mine presents a unique opportunity to measure the impact of slurry temperature on process performance under full scale plant, conditions. This thesis analyzes an energy balance approach to model the seasonal variations in slurry temperature throughout the Fort Knox mill. The mill utilizes both gravity concentration and cyanidation for gold recovery. Models were developed to accurately predict the impact of slurry temperature on cyanide leach, carbon adsorption and cyanide destruction kinetics. The energy balance model, combined with the kinetics models, was used to accurately predict the gold recovery and subsequently to justify the installation of a tailings wash thickener to recovery heat from the mill tailings. A substantial portion of this thesis is dedicated to the development of these models, analysis of the post expansion plant performance, and summarizing project economics. Gold in the Fort Knox deposit is generally less than 100 microns in size and contained in quartz veins and along shears within the host granite, at an average gold grade of 0.8 g/metric ton. In April 2001, the mill began processing ore from a satellite ore deposit, the True North Mine, as a blend with Fort Knox ore. The gold grade in the True North deposit averages 1.5 g/metric ton and can be associated with pyrite, arsenopyrite and stibnite. An unexpected drop in gold recovery resulted from processing the blended ore and was the subject of an extensive laboratory evaluation. Laboratory results suggested that the leach kinetics of the coarse gold particles were significantly impacted, when the blended ore was processed, and that the impact could be reduced, or eliminated, with the addition of lead nitrate. Subsequently, a lead nitrate addition scheme was implemented at the Fort Knox mill. A portion of this thesis is dedicated to a review of the laboratory program, an evaluation of the environmental impacts and a summary of plant performance, when utilizing lead nitrate at the Fort Knox Mine.
-
Gis-Based Approaches To Slope Stability Analysis And Earthquake -Induced Landslide Hazard ZonationThis dissertation presents newly developed GIS-based deterministic and probabilistic approaches for slope stability analysis and earthquake-induced landslide hazard zonation. The described approaches combine numerical slope stability analysis with GIS spatial analysis to evaluate earthquake-induced slope failures, both shallow and deep-seated. The study has four major research components. The first component is a GIS-based procedure which was developed based on one-, two-, and three-dimensional (1D, 2D, and 3D) deterministic approaches to slope stability analysis and landslide hazard zonation. Slope stability methods in the GIS-based procedure included the infinite slope model, the block sliding model, the ordinary method of slices, the Bishop simplified method, and the Hovland's column method. The second component focuses on causative factors analysis of earthquake-induced landslide hazards. This component also discusses the determination of peak ground acceleration for slope stability analysis. The third component consists of an evaluation of the topographic effect of ground motion and the seismic response in the Balsamo Ridge area in Nueva San Salvador. The fourth component is concerned with the regional and site-specific landslide hazard zonation, using newly developed models for landslide hazard assessment in Nueva San Salvador. The slope stability and landslide susceptibility were mapped in terms of slope stability index (factor of safety, critical acceleration, Newmark displacement, failure probability, and reliability index). The landslides triggered by an earthquake on January 13, 2001 in El Salvador provide a setting for the calibration of results from GIS-based approaches. The procedures developed in this research proved to be feasible and cost-effective for slope stability analysis and earthquake-induced landslide hazard zonation.
-
Predictive Performance Of Machine Learning Algorithms For Ore Reserve Estimation In Sparse And Imprecise DataTraditional geostatistical estimation techniques have been used predominantly in the mining industry for the purpose of ore reserve estimation. Determination of mineral reserve has always posed considerable challenge to mining engineers due to geological complexities that are generally associated with the phenomenon of ore body formation. Considerable research over the years has resulted in the development of a number of state-of-the-art methods for the task of predictive spatial mapping such as ore reserve estimation. Recent advances in the use of the machine learning algorithms (MLA) have provided a new approach to solve the age-old problem. Therefore, this thesis is focused on the use of two MLA, viz. the neural network (NN) and support vector machine (SVM), for the purpose of ore reserve estimation. Application of the MLA have been elaborated with two complex drill hole datasets. The first dataset is a placer gold drill hole data characterized by high degree of spatial variability, sparseness and noise while the second dataset is obtained from a continuous lode deposit. The application and success of the models developed using these MLA for the purpose of ore reserve estimation depends to a large extent on the data subsets on which they are trained and subsequently on the selection of the appropriate model parameters. The model data subsets obtained by random data division are not desirable in sparse data conditions as it usually results in statistically dissimilar subsets, thereby reducing their applicability. Therefore, an ideal technique for data subdivision has been suggested in the thesis. Additionally, issues pertaining to the optimum model development have also been discussed. To investigate the accuracy and the applicability of the MLA for ore reserve estimation, their generalization ability was compared with the geostatistical ordinary kriging (OK) method. The analysis of Mean Square Error (MSE), Mean Absolute Error (MAE), Mean Error (ME) and the coefficient of determination (R2) as the indices of the model performance indicated that they may significantly improve the predictive ability and thereby reduce the inherent risk in ore reserve estimation.
-
Analysis and evaluation of fragment size distributions in rock blasting at the Erdenet MineRock blasting is one of the most important operations in mining. It significantly affects the subsequent comminution processes and, therefore, is critical to successful mining productions. In this study, for the evaluation of the blasting performance at the Erdenet Mine, we analyzed rock fragment size distributions with the digital image processing method. The uniformities of rock fragments and the mean fragment sizes were determined and applied in the Kuz-Ram model. Statistical prediction models were also developed based on the field measured parameters. The results were compared with the Kuz-Ram model predictions and the digital image processing measurements. A total of twenty-eight images from eleven blasting patterns were processed, and rock size distributions were determined by Split-Desktop program in this study. Based on the rock mass and explosive properties and the blasting parameters, the rock fragment size distributions were also determined with the Kuz-Ram model and compared with the measurements by digital image processing. Furthermore, in order to improve the prediction of rock fragment size distributions at the mine, regression analyses were conducted and statistical models w ere developed for the estimation of the uniformity and characteristic size. The results indicated that there were discrepancies between the digital image measurements and those estimated by the Kuz-Ram model. The uniformity indices of image processing measurements varied from 0.76 to 1.90, while those estimate by the Kuz-Ram model were from 1.07 to 1.13. The mean fragment size of the Kuz-Ram model prediction was 97.59% greater than the mean fragment size of the image processing. The multivariate nonlinear regression analyses conducted in this study indicated that rock uniaxial compressive strength and elastic modulus, explosive energy input in the blasting, bench height to burden ratio and blast area per hole were significant predictor variables in determining the fragment characteristic size and the uniformity index. The regression models developed based on the above predictor variables showed much closer agreement with the measurements.
-
Neural-Network Modeling Of Placer Ore Grade Spatial VariabilityTraditional geostatistical methods have been used in ore reserve estimation for decades. Research in the last two decades or so has added a number of other statistical methodologies for ore reserve estimation procedures. Recent advances in neural networks have provided a new approach to solve this problem. This thesis is focused on the Neural-network modeling for the estimation of placer ore reserve. Due to the spatial variability, multiple dimensional inputs and very noisy drill hole sample data from the selected region, it requires that the neural-network be organized in a multiple-layers to handle the non-linearity and hidden slabs for smoothing the predicted results. Various neural-network architectures are investigated and the Back-propagation is selected for modeling the ore reserve estimation problem. Sensitivity analysis is performed for the following parameters: the type of neural-network architecture, number of hidden layers and hidden neurons, type of activation functions, learning rate and momentum factors, input pattern schedule, weight updated, and so on. The influences of these parameters on the predicted output are analyzed in details and the optimal parameters are determined. To investigate the accuracy and promise of neural network modeling as a tool for ore reserve estimation, the ore grade and tonnage of Neural-network output is compared with those estimated by geostatistical methods under various cut-off grades. In addition, the overall performance is also validated by the analysis of R-squared (R2), Root-Mean-Squared (RMS), and the comparison between predicted values and 'actual' values. As the final part of this study, the optimized Neural Network was used to estimate the distribution of placer gold grade and volume of gold resource in offshore Nome. The predicted results for all the mining blocks in the lease area are validated by checking the values of RMS, R2, and Scatter plots. The estimated gold grades are also presented as contour maps for visualization.
-
Two dimensional computational fluid dynamics model of pollutant transport in an open pit mine under Arctic inversionA better understanding of the microscale meteorology of deep, open pit mines is important for mineral exploitation in arctic and subarctic regions. During strong temperature inversions in the atmospheric boundary layer--which are common in arctic regions during the winter--the concentrations of gaseous pollutants in open pit mines can reach dangerous levels. In this research, a two dimensional computational fluid dynamics (CFD) model was used to study the atmosphere of an open pit mine. The natural airflow patterns in an open pit mine are strongly dependent on the geometry of the mine. Generally, mechanical turbulence created by the mine topography results in a recirculatory region at the bottom of the mine that is detached from the freestream. The presence of a temperature inversion further inhibits natural ventilation in open pit mines, and the air can quickly become contaminated if a source of pollution is present. Several different exhaust fan configurations were modeled to see if the pollution problem could be mitigated. The two dimensional model suggests that mitigation is possible, but the large quantity of ventilating air required would most likely beimpractical in an industrial setting.
-
Geodatabase development and GIS based analysis for resource assessment of placer platinum in the offshore region of Goodnews Bay, AlaskaGoodnews Bay, southwest Alaska, is known for extensive Pt reserves that have their source in the neighboring Red Mountain. The reserves potentially extend offshore into the Bering Sea. This study aims at developing a geodatabase to integrate all offshore platinum related data collected by researchers and agencies in the past, with the intent to identify data gaps. Based on these data gaps 49 new areas were sampled for Pt and geophysical data were collected in summer 2005. Spatial distribution map for offshore Pt was created using a new Multiple Regression Pattern Recognition Technique (MRPRT) that gave an R²=0.76, a significant improvement from standard GIS based geospatial techniques. Four potential Pt exploration areas were delineated, including one area where drowned ultramafics and buried alluvial channels co-occur. Coastal currents influenced the surficial platinum accumulations, and no clear relation between Pt distribution and sand bars in the far offshore could be established.
-
Automatic detection of sensor calibration errors in mining industrySensor errors cost the mining industry millions of dollars in losses each year. Unlike gross errors, "calibration errors" are subtle, develop over time, and are difficult to identify. Economic losses start accumulating even when errors are small. Therefore, the aim of this research was to develop methods to identify calibration errors well before they become obvious. The goal in this research was to detect errors at a bias as low as 2% in magnitude. The innovative strategy developed relied on relationships between a variety of sensors to detect when a given sensor started to stray. Sensors in a carbon stripping circuit at a gold processing facility (Pogo Mine) in Alaska were chosen for the study. The results from the initial application of classical statistical methods like correlation, aggregation and principal component analysis (PCA), and the signal processing methods (FFT), to find bias (±10%) in "feed" sensor data from a semi-autogenous (SAG) grinding mill operation (Fort Knox mine, Alaska) were not promising due to the non-linear and non-stationary nature of the process characteristics. Therefore, those techniques were replaced with some innovative data mining techniques when the focus shifted to Pogo Mine, where the task was to detect calibration errors in strip vessel temperature sensors in the carbon stripping circuit. The new techniques used data from two strip vessel temperature sensors (S1 and S2), four heat exchanger related temperature sensors (H1 through H4), barren flow sensor (BARNFL) and a glycol flow sensor (GLYFL). These eight sensors were deemed to be part of the same process. To detect when the calibration of one of the strip vessel temperature sensors, S1, started to stray, tests were designed to detect changes in relationship between the eight temperature sensors. Data was filtered ("threshold") based on process characteristics prior to being used in tests. The tests combined basic concepts such as moving windows of time, ratios (ratio of one sensor data to data from a set of sensors), tracking of maximum values, etc. Error was triggered when certain rules were violated. A 2% error was randomly introduced into one of the two strip vessel temperature data streams to simulate calibration errors. Some tests were less effective than others at detecting the simulated errors. The tests that used GLYFL and BARNFL were not very effective. On the other hand, the tests that used total "Heat" of all the heat exchanger sensors were very effective. When the tests were administered together ("Combined test"), they have a high success rate (95%) in terms of True alarms, i.e., tests detecting bias after it is introduced. In those True alarms, for 75% of the cases, the introduction of the error was detected within 39.5 days. A -2% random error was detected with a similar success rate.
-
Experimental and numerical simulation of hydraulic fracturingHydraulic Fracturing (HF) has many applications in different fields such as stimulation of oil and gas reservoirs, in situ stress measurements, stress relief for tunneling projects as well as in underground mining applications such as block caving mining. In the HF process, high pressure fluid is injected into a well to generate fractures in tight rock formations. This technique is particularly suitable for developing hydrocarbon energy resources in tight rock formations such as shale with very low permeability. An experimental setup was designed and developed to simulate the HF process in the laboratory. Cubic plaster specimens were molded and HF experiments were conducted with simulated plaster models. Five laboratory tests were performed on cubic specimens under different stress conditions. Because the uniaxial compressive strength of the plaster was about 1600 psi, in all experiments the applied vertical stress was 1000 psi to avoid breaking the specimens before injection of fluid. The differential horizontal stress varied from 100 to 500 psi. These stress levels are related to shallow formations in a real environment. It was observed that increasing the differential horizontal stress by 100 psi, the minimum pressure required to initiate HF decreases about 100 psi. These results were in agreement by 2D failure criterion of HF. All in all, the small scale HF experiments were conducted successfully in the rock mechanics lab. It was observed that vertical hydraulic fractures would propagate along maximum horizontal stress, which is in agreement with propagation of HF theory. Three-dimensional (3D) numerical models were developed and computer simulations were conducted with ABAQUS, a commercially available finite element analysis (FEA) software. The numerical simulation results compared favorably with those from the laboratory experiments, and verification and analysis were carried out. Since the results obtained from the numerical model were in agreement with the results of experiments and verified the correctness of the model, further investigation was carried out with developed computer models. Several scenarios with different vertical stresses and different levels of horizontal stress were simulated. A statistical software, R, was used to generate a 3D failure criterion for the HF in shallow formations.... It can be stated that in shallow formations, vertical stress has the least effect among stress components on the minimum pressure required to initiate HF.
