Search Documents

By I Canbulat, J Emery, L Yao and C Zhang

This paper documents a process of data analytics and machine learning applied to coalmine roof convergence data. It follows a publication documenting the formation of the database. For each set of analyses Python scripts were utilised to extract data from an SQL database, with subsequent analyses also executed in Python, utilising open-source tools such as scikit-learn. Preliminary analyses with classic machine learning algorithms included multiple logistic regression, K-nearest neighbours, decision tree, random forest, support vector machine, kernel support vector machine, artificial neural network (ANN), naïve Bayes and multiple linear regression. Results indicated that relationships were complex and non-linear, with the artificial neural network most suitable. An ANN model was then refined through a standard process of hyperparameter optimisation and data augmentation to arrive at a final model. Due to the black box nature of the ANN further insight into how the parameters interacted was sought and found with Shapley additive explanations (SHAP values). SHAP values utilise Game Theory, where each input parameter or feature is a ‘player’ while the data set is the ‘team’. The SHAP value is the impact of each player on the target value, essentially establishing the contribution of each input parameter. The results from all analyses are delivered in the paper, together with insights, applications, strengths, and limitations of the methodology.

Nov 29, 2022

By Marshall T. E, Finfinger G. L

Gulf Coast domal salt mining operations have had sporadic encounters with methane gas emitted without any warning. In virtually all confirmed cases, the methane was localized in methane enriched zones. In a few cases large quantities of methane were expelled rapidly with salt outbursts, resulting in dangerous methane concentrations. Estimates of methane liberated by face blasts in methane-enriched areas is needed in determining how much air should be supplied to the face for proper dilution. A relationship between the gas content of salt samples collected from the face and the amount of methane liberated would be helpful in estimating how much methane will be produced by a face blast. Little is known about the layering or stratification of methane in large area-low velocity airways. This publication reports on three tests conducted by the Bureau of Mines which monitored the immediate return air of four faces during blasting. The test locations were in a methane gas-enriched section that had previously experienced two large outbursts and many smaller ones. A comparison was made between gas content values of face salt samples from the dissolution test and the gas liberation quantities. The characteristic movements of the methane in a large room with low air velocities were charted. This will help to better understand the factors involved in proper placement of methane sensors as a warning system.

Jan 1, 1988

By Onal G, Celik MS, Dogan MZ

Chromite tailings assaying 21.2 per cent Cr2O3 was upgraded to 48.7 per cent Cr2O3 at a recovery of 72.0 per cent using column flotation with a mixture of anionic collectors at pH 11. Electrokinetic methods were used to elucidate the mechanism of collector adsorption and correlate the findings with the optimum flotation conditions. Multivalent constituent ions such as Cr3+, A13+, Fe2+ and Mg2+ were found to adsorb on chromite surface and shift the zpc of chromite in the direction of oxide form of the mineral. The anionic collectors also imparted a negative charge to chromite in the entire pH region. The higher zeta potentials obtained upon adding the anionic collectors is ascribed to the adsorption of anionic surfactant onto the positively charged chromite surface. The positive sites are generated through formation of significant amount of hydroxy complexes of magnesium at pH 11. Complementary infrared data reveal the absence of a shift in the spectra indicating the possibility of physical adsorption in the system.

Jan 1, 1993

The story of the rise, progress, and development of our modern railway system is a long but an exceedingly interesting one. The phases it has passed through, the discussions that have arisen from time to time on the most economic modes of construction and working, and the final result these various discussions have led to, are full of importance and interest to the engineer. It is not my intention to deal historically with the subject. I desire simply to speak to the requirements of the hour in Australia. It will not be disputed, I think, that these colonies have reached a point in railway construction and management when, if additional lines to any extent are still to be built, a different principle must be adopted. There are yet large tracts and areas of country to be tapped by the rail and the locomotive, which for years to come must remain untouched if the same expensive gauges and modes of construction are continued. We have yet to see clearly the necessity of introducing a system of light lines that will pay their own way and at the same time augment the traffic of our main lines. This question really no longer remains an unsolved problem. It lies, in a few words, in the acceptance of the, narrowest gauge and the simplest mode of construction that can constitute a pacticable railway. I fear it is hopeless to suggest that private companies should be legally permitted to construct such lines. It seems as if it were next to impossible to move official authorities to enter upon the construction of lines on this principle. It is, all the same an undoubted fact that if private companies could obtain Acts of Parliament, that in several of the colonies a new impetus would be given to railway construction.I may point out that not only has the principle of narrow gauge light railways been adopted in India but private companies are obtaining power to build them. Outside of :England many narrow gauge lines have been made, and in France, Denmark, Italy, and Sweden they have been highly successful.

By C A. Brannon, C A. Dipas, H A. Vega, D Villa

The development and construction of the Grasberg block cave mine (GBC) is in progress, currently targeting initiation of caving in late 2018. Current production planning anticipates ramping up to full production of 130 000 to 160 000 t/d by 2027. As GBC approaches the production phase, there is a high priority on increasing the accuracy and level of detail in the production plan. The advance of the cave front directly controls the drawbell opening rate that is achievable. Therefore, an essential component of this plan is to produce an undercutting and drawbell sequence that honours the caving parameters that have been set for the mining blocks by the GBC production planning team.The PT Freeport Indonesia (PTFI) mine planning team and the Geovia caving business unit have developed a computational tool (Undercut Tool) capable of generating an undercutting sequence that follows a set of caving rules, such as advance rate per tunnel, maximum blasted area per period, control of lead-lag and shape of the undercut face. This tool also produces a drawbell sequence that honours the undercutting advance sequence, which is ultimately used to generate a life-of-mine production plan. This paper describes the inputs and outputs of the Undercut Tool that have been used to compute the long-term production plan for the GBC mine. It will further describe how the Undercut Tool takes into account operational constraints and geotechnical considerations that must be considered when developing a defendable production plan. Included is a discussion of the initial results of the planning outcomes from the tool; it will demonstrate how this work will support specific aspects of the plan such as the detailed undercut production drill and blast plans for the drill ring sequence, as well as how that plan will drive the drawpoint opening plan.CITATION:Villa, D, Brannon, C A, Dipas, C A and Vega, H A, 2017. Integrated drawbell and undercut scheduling using a computational tool at the Grasberg block cave mine, in Proceedings 13th AusIMM Underground Operators’ Conference 2017, pp 271–276 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Oct 16, 2017

By Agrawal R. D, Kapoor M. L

Activity of FeO in the ternary system MgO-FeO-CaO in contact with metallic iron has been measured at 1000 K employing the cell.PtIFe, (Mg, Fe, Ca)OIZr02 - CaOIFe, FeOlPt. Measured data are used to study the phase stability relations in the system MgO-FeO-CaO and an isothermal section at 1000 K through this system is drawn.

Jan 1, 1987

By K McSweeney, A Marton

This paper discusses the development of a strategy for ensuring long-term geotechnical stability of final pit walls at AngloGold AustraliaÆs Sunrise Dam Gold Mine. The strategy makes use of improved blasting techniques, changes to the mining sequence and other geotechnical techniques. Only the blasting techniques will be discussed in this paper. These were developed from numerous trials over a period of several months from May 2003 through to November 2003. The focus was to reduce crest loss damage and thereby improve final wall integrity, maintain designed safe berm widths, and ensure long-term stability. Crest loss mechanisms are now understood and are explained in this paper. The preferred solution is outlined; which is a combination of improved blasting and changes in the mining sequence to cater for pre-existing ground conditions. These ground conditions include dynamic hyper saline water with sometimes intersecting steeply dipping discrete structures. Project requirements were defined as mitigating or eliminating crest loss using 127 mm diameter pre-splitting and targeting minimal production blasting vibration levels against the walls. Blast results were measured by post blast visual inspections and video. Blast design input utilised the predictive algorithms built into simulation software when attempting to minimise vibration levels against the walls. Protection of the already fired pre-split line is now achieved using small charges against the wall and timing sequences which form a mask for the later firing main charges. The pre-split spacing is now modified to suit the geology. There eight blast domains identified around the perimeter of the pit, each having varying joint orientation. In addition, blasting the bench adjacent to the berm crest has particular considerations not required by the middle and lower benches. The preferred solution was to purposefully translate this damaged top section of the pre-split bench into a planned subsequent dig zone. The design of this top æweightÆ is a direct response to the nature of the crest loss mechanism.

Jan 1, 2004

The Channel Tunnel is an outstanding example of an undersea tunnel and its design illustrates most of the unique features of this type of tunnel. The marine site investigation was partic- ularly vulnerable to adverse weather conditions. To minimise this interference the time the drilling vessel was required to remain on station was kept to a minimum by expediting the drilling and the insitu tests as far as was practicable. To avoid additional unforeseen costs the contract documents were framed to take particular account of marine practice. The limited access at the coast line meant that for economy, maximum tunnel advance rates had to be attained. Ground excavation, handling and disposal methods and the provision of the lining were optimised and equipment specified in detail to ensure that no one com- ponent was delayed by a lack of capacity or mechanical failure in any other.

Jan 1, 1976

Samples weighing 0·5 g are fused with sodium hydrogen sulphate, leached with 5 ml. of 4 M nitric acid and diluted to 10 ml. with water. An aliquot of the clear solution is boiled with concentrated nitric acid and then 2 ml. of a phosphotungstate solution are added. After dilution to 10 ml. with water, the solution is heated in a boiling water-bath for not less than 10 min. When cold, the solution is extracted with 2 ml. of iso-amyl alcohol and the yellow phosphotungstovanadate complex is compared with a standard series. Concentrations down to 4 p.p.m. may be analysed at a rate of 100 samples per man-day of 8 hr.

Jan 1, 1971

By M J. Nimmo

extremely powerful tools for building predictive and generative models. ML can be used for building highly accurate regression and classification models. But without careful data science and statistical analysis, a regression or classification model could be highly biased and completely wrong – and we may not even realise. An example rock density estimation task will be used to illustrate the potential gains and possible pitfalls in using ML and AI in Mineral Resource estimation. What could possibly go wrong? Nothing! For most simple regression or classification problems, very little. For density estimation, using the global mean may be all that is possible, but the resource classification would need to reflect the uncertainty in global estimation of tonnage. Everything! From data collection and measurement, from selecting ML and AI techniques and designing the analysis workflow to our own cognitive bias. The things that could go wrong are vast and may include but not limited to asking the wrong question, not asking questions, data bias, measurement errors, insufficient data coverage, not collecting the right variables, not collecting enough observations, collecting clustered samples, unbalanced data, filtering out data without reason (outliers), clipping data (outliers), relying on automatic feature selection, including irrelevant variables, excluding relevant variables, geological interpretation, changing context, blindly following best practice, splitting small data sets, subsetting large data sets, using the wrong tool for the task, insufficient budget to complete the analysis, not allowing enough time for testing and experimentation (the science), focusing too much on building the model (the engineering), not learning from the data, our assumptions, our preconceptions, and our skill. This paper will explore the question of how training regression models to predict rock density are affected by gaps in the data (missing observations and missing variables). What could possibly go wrong?

May 24, 2023

The segment of the Upper Cenozoic volcanic chain (CVZ) located on the western border of South America (16-28¦S lat.) is constituted by an association of strato- volcanoes and ignimbrite sheets, with high K calco-alkaline affinities. Strata-bound volcanoes, mainly of andesitic and dacitic composition, present along the whole chain, and various hydrothermal alteration zones, on the slopes and in the cores of eroded volcanic structures. Some of these altera- tion zones bear epithermal precious metal ores and sulphur deposits, which at present are being exploited. The studied Salar de Gorbea area, located in the Andes of northern Chile (25-26¦S lat.) contains a numer- ous group of Miocene to Quaternary stratovolcanoes, with advanced argillic and extreme silicification hydrothermal alteration in their cores and on their slopes, favoring the formation of great volumes of alunite within altered zones and sulphur deposits in the higher parts of the volcanic edifices. Simultaneously, siliceous hydrothermal breccia of the phreato-magmatic type, with evidences of hydromag- matic explosions are recognizable, allowing, in some cases, the ascent towards the surface of rock fragments belonging to the Paleozoic pre volcanic basement. Radiometric ages obtained in some of the hydro- thermal alteration zones (K-Ar in alunite) are slighty young- er than the unaltered rocks which constitute the strato- volcano. Trace element (Hg, As, Sb, S, Au, Ag, Cu, Pb, Zn, Mo) patterns in orientative rock samplings show that the studied alterations zones correspond to high sectors of the cpithermal system, without evidences of metallic minerali- zation on surface. The weak erosive action due to the ex- tremely and climatic conditions in this segment of the Andes, has not allowed the cropping out of deeper areas of the hydrothermal alteration zones, in contrast to other Miocene more eroded volcanic edifices and Miocene vol- canic terranes of the Andean Range.

Jan 1, 1987

By D Adhikary, M Khanal, B Poulsen, H Guo

During longwall mining, the ability to reliably estimate the height of longwall mining-induced fractures is critically important in order to effectively manage mining-induced water inflows and aquifer interference. The extent of mining-induced subsurface deformation, fracture, surface subsidence and aquifer interference is largely controlled by local lithological conditions, mining methods and mine layouts. Therefore for the purpose of quantifying mining impacts it is imperative to incorporate the local geology and mine plan parameters in a mine scale model. Two different numerical models based on 3D continuum (Finite Element Method) and 2D discrete (Discrete element method) domains, respectively, are developed to incorporate the mine site parameters of a study site in Australia. Both the models are calibrated and validated with measured mine data. The continuum numerical modelling method is used to investigate the change in permeability of the overburden strata due to underground extraction considering the full 3D nature of geology and mine layout. The discrete element modelling method is used to simulate the initiation and propagation of fractures, allowing the caving of overburden strata to be represented in a realistic manner, and the permeability enhancement of the strata overlying the goaf to be calculated from first principles. Permeability enhancement in the discrete element model is a direct function of the fractures induced in the deformation/caving process; fracture location, extent, aperture and connectedness. From the numerical modelling under the simulated condition for the mine studied here, it is predicted that the permeability of the caving zone from the goaf to five times the mining height increases by almost six orders of magnitude. In the fracture zone to approximately ten times the mining height, the change in permeability is predicted to be around two to three orders of magnitude. At the height of more than ten times the mining height the magnitude of permeability increased is predicted to be one to two orders of magnitude. CITATION:Khanal, M, Poulsen, B, Adhikary, D and Guo, H, 2018. Modelling of height of connective fractures during longwall mining, in Proceedings The Fourth Australasian Ground Control in Mining Conference (AusRock), pp 254–265 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Nov 30, 2018

By J Janus, J Krawczyk

Local accumulation of methane in workings of a longwall is one of the most common causes of mining accidents. This gas has many sources: it may leak from goaf through permeable sealings, or through gaps between roof support sections. This stream is then mixed with ventilation air. In the case of auxiliary ventilation ducts, or free jet fans, we have to consider the propagation of a clean air stream in areas with an air-methane mixture. Therefore, modelling the mixing process of methane with ventilation air is important for determining the hazardous areas and thus to ensure safety. Computational fluid dynamics (CFD), specifically the finite volume method, provides tools for such modelling. However, this method requires validation. For this purpose, we have compared measurements with simulations for selected test cases having sufficient similarity to the present problems. In typical methane explosion hazard areas, such as a longwall, it is difficult to make measurements sufficiently to confirm credibility of 3D CFD simulations. As a relevant alternative, an injector methane drainage station placed in an underground gallery has been selected. The stream of drained goaf methane is entrained by compressed air powered injectors and then mixed with ventilation air. Research in this area does not interfere with mining operations, it can be performed safely in a repeatable way. Flow within the site has many similarities to conditions in mine drifts of the longwall area. Comparative research of measuring the methane concentration distribution in the mine drift downstream injector station, with simulations of propagation of the methane stream, was carried out. For measuring the methane concentration distribution in the mine drift a multipoint measurement velocity system developed at the Strata Mechanics Research Institute was used. For the simulation a finite volume method for turbulent flow of gas mixtures was applied. Preliminary results of comparison are shown as graphs and tables.CITATION:Janus, J and Krawczyk, J, 2013. An analysis of the mixing of air and methane in the stream produced by the mine injector station - Presenting preliminary results of measurements and modelling, in Proceedings The Australian Mine Ventilation Conference , pp 49-56 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Jul 1, 2013

By J Poniewierski, Varcos D

Efficient blasting requires constant vigilance with respect to the actual layout being used in the field. Objective measurement techniques are required in laying out a pattern as the 'by eye' or 'thumb suck' methods cannot guarantee consistent results in layout. Erratic blast results are the consequence of such an approach. In a case study at Sons of Gwalia mine, it was demonstrated that incorrect front burdens had a snowballing effect on otherwise carefully laid out blasts, resulting in both excessive amounts of boulders caused by preconditioning of bench fronts and tops, and tight digging. Both problems caused lower excavator productivity and excessive excavator damage, as well as making it difficult to set out subsequent blasts correctly. Several field measurement methods are available to aid in determining correct blast layouts and one is described in this paper.

Jan 1, 1994

By S Nkwanyana, I Govender

Batch experiments in a 0.6 m diameter mill have shown that improvements of about 13 per cent in the efficiency of energy utilisation can be realised in ball mills by replacing 25 per cent of the ball charge by an equivalent volume of cost-free critical size pebbles from AG/SAG mills. For a mill of a given size, speed and charge filling, the rate of size reduction of ore particles depends on (i) the frequency of grinding media collisions (ii) the probability of capture of ore particles between the colliding grinding media (iii) the probability of fracture of the captured particles. The probability of fracture depends on the magnitude of collision energy. The distribution of collision energy into shear and impact events determines the relative strengths of attrition breakage and impact breakage. The efficiency of energy utilisation depends on which of the two mechanisms of breakage is more desirable in a particular milling stage. In this paper collision data from a series of DEM simulations of a mill charged with steel media and a mixture of steel media and pebbles is presented in energy spectra and spatial distribution maps. The data provides the necessary detail to elucidate the improvements in the efficiency of energy utilisation in ball mills when using a mixture of pebbles and steel media.

Aug 24, 2022

By Giles D, Wegmann D

The Mt Todd Gold Project is located 42 km northwest of the township of Katherine within the Northern Territory of Australia. It is situated in the southeastern portion of the Pine Creek Geosyncline. The Batman gold deposit, the largest of several deposits at Mt Todd, hosts an identified mineral resource of 92.9 million tonnes (Mt) grading 1.4 grams/tonne (g/t) as at 31 December 1993. Host rocks at Batman are assigned to the Early Proterozoic Burrell Creek Formation, comprising greywacke and siltstone interbedded with shale. Gold mineralisation is associated with a very broad zone of sheeted and stockwork quartz-sulphide veins that occupy a corridor of intense brittle fracture and faulting. A conceptual model of formation involves the dilation of joints and micro-faults within overlying sediments during sinistral movement along a basement shear zone. Synchronous granite intrusion generated convective circulation of crustal fluids through the sedimentary pile resulting in the formation of auriferous quartz veins within favourable structures. The Batman zone of mineralisation is not directly related to fold structures which typify several other major gold deposits within the Pine Creek Geosyncline.

Jan 1, 1994

By H Jalalifar, N Aziz

In the ten years since the University of NSW proposed a pillar design methodology for bord and pillar operations, the Australian coal mining industry has changed substantially. What was primarily a bord and pillar design approach is now being applied to chain pillar design in longwall mines where the requirements are substantially different. The dimensions of chain pillars can impact on tailgate conditions (roof, rib and floor), seal performance, and surface subsidence. The status of chain pillar design practice in Australia is reviewed, with a focus on defining pillar strength, chain pillar loading, and assessing the performance of the roof/pillar/system. A new pillar strength equation is proposed for Australian coal that applies for all width/height ratios. An alternative analysis of probability of failure of chain pillars is presented.

Jan 1, 2005

By A Gerson, S Miller, R Smart, J Thomas

Environmental Geochemistry International (EGi) and the Ian Wark Research Institute (IWRI) collaborated in a three-year Australian Minerals Industry Research Association (AMIRA) research project (Project Number P387A) focussed on the prediction and kinetic control of acid mine drainage. As part of this project, single addition, and sequential net acid generation (NAG) testing was carried out on sulfide mineral samples. The test samples comprised a range of synthetic waste materials prepared using a quartz bulk and controlled quantities of the commonly occurring sulfide minerals; pyrite (FeS2), pyrrhotite (Fe7S8), arsenopyrite (AsFeS), chalcopyrite (CuFeS2), chalcocite (Cu2S), covellite (CuS), bornite (Cu5FeS4), galena (PbS) and sphalerite (ZnS). NAG test results clearly show that there are significant differences between the acid forming potential of sulfide minerals. The NAG test results for each mineral were compared with theoretical acidity values (NAPP*). The NAPP* calculation was based on stoichiometric oxidation reactions assumed for each of the sulfide minerals. Sequential NAG values for pyrite corresponds closely to predicted NAPP* values, confirming the assumption that pyrite oxidation occurs according to the stoichiometry assumed. For the pyrite sample with 5 wt per cent S, two NAG stages were sufficient for all of the sulfur to be released to solution. Since sulfides in waste materials will generally be dominated by pyrite, results also confirm the validity of including NAG testing for ARD evaluation of most waste rock. NAG testing confirms that of the sulfide minerals examined, only pyrite, pyrrhotite, arsenopyrite and chalcopyrite are acid producing, ie able to produce leachate with a pH less than 4.5. In the case of sphalerite, galena, bornite, chalcocite and covellite the pH of the final NAG solution was greater than 4.5. The significant differences between the acid forming potential of different sulfide minerals could lead to overestimation of acid producing potentials based on NAPP values calculated from total per cent S alone, ie assuming all S is present as pyrite. The oxidation of galena during the NAG test resulted in the consumption of acid. The examination of the galena surface by X-ray photoelectron spectroscopy (XPS) after oxidation in peroxide solution indicated the presence of polysulfide. The reaction of sulfide to polysulfide is acid consuming, thus explaining the net acid consuming behaviour of galena in the NAG test. This reaction pathway may provide a possible explanation for the lower than expected acid production observed for some of the other sulfide minerals examined.

Jan 1, 2003

By M F. L Gibson

The issue of flyrock is very important in the operation of all mineral extraction sites that use blasting. Unlike the other environmental effects of blasting, ground vibrations and air overpressure, flyrock is (or should be) a rare event, with severe consequences which may include damage to property, injury or fatalities. These physical consequences are often matched by a loss in public confidence in the management of the operation, mistrust and bad press. In this work a deterministic analysis of flyrock was carried out using the equations of a particle in motion with drag. The solutions to these equations were used to reproduce the curves of Lundborg et al (1975) for the relationship between flyrock boulder size and maximum ejection distance. A probabilistic approach based on Monte Carlo simulation was built up with the analytical model using published data together with local case study material. A typical quarry blast situation was modelled and estimates of safe standoff distances were determined using risk assessment techniques.

Jan 1, 2001

By L Alem

In the field of mineral extraction, complex technologies are being introduced in mines, operators/technicians rely on assistance from an expert (or more) in order to keep their machines functioning. Personnel with such expertise, however, are not always physically on-site. There is a growing need for innovative technological solutions to support effective delivery, just in time, of remote guidance between an offsite expert and an on-site maintenance operator. To meet this need, we have, as part of our Human System Integration project within the CSIRO’s Minerals Down Under (MDU), Transforming the Future Mine theme, explored the usefulness of a number of emerging technologies including: augmented reality, wearable systems and touch table interfaces. This paper presents ReMoTe, an augmented realty system for remote guiding using a touch table interface for the expert station and a combination of augmented reality and a wearable system at the operator end. In designing this system we were interested in exploring the extent to which it can increase the productivity and safety of maintenance personnel. This paper also presents Mobile ReMote, our recent development of a fully mobile remote guiding system. In Mobile ReMote, both the expert and the operator are using a near eye display and wearing a light computer in a back pack. The expert is able to guide the operator using his/her hands on the virtual display, while the operator is able to see the instructions on the rear eye display. Workers in the mining industry are highly mobile, a mobile expert and operator station will meet this need.

Nov 22, 2011