Forecasting time-to-failure of machine using hybrid Neuro-genetic algorithm – a case study in mining machinery

Forecasting of time-to-failure is an important aspect of a mining machine for the performance assessment, fault detection and schedule maintenance. The knowledge of failure time allows more defined arrangement of preventive maintenance. Traditional methods, including lifetime distribution models, fault tree analysis and Markov models, have a limitation of assuming a specific statistical distribution function to fit the failure time data. In this study, a hybrid data-driven method using neural network and genetic algorithm is proposed to forecast failure time. The forecasting model was developed using neural network algorithm and all the neural network parameters, i.e. input nodes, hidden nodes and the learning algorithms, are selected automatically using the genetic algorithm. The developed model was validated using the failure data of a mining machine. A case study was conducted investigating a load-haul-dump machine (LHD) in the mining industry. Failure historical data for the LHD machine were collected, and cumulative failure time was calculated for time-to-failure forecasting. Study results demonstrate that the developed model performs satisfactory in the prediction of next failure time. A comparative study reveals that the proposed method performs better than existing methods.     

A GIS-based methodology for identifying sustainability conflict areas in mine design – a case study from a surface coal mine in the USA

Through proper design, coal mining operations can contribute to the social, economic and environmental aspects of sustainable development. The regulatory and permitting programmes for coal mining in the United States, which often focus on several largely environmental parameters, are barriers to holistic consideration of these three sustainability pillars. Some changes in the current US regulatory framework may be necessary to allow for systematic consideration of economic, environmental and social factors in order to achieve more sustainable development of mineral and energy resources. In the quest towards more sustainable mining activities, it has been suggested that mine design may be optimised in the context of sustainability using a systems engineering approach that simultaneously considers economic, environmental and social factors. The use of geographic information system (GIS) tools may aid in this approach by allowing for spatial analyses of various resources (e.g. mineral values, water resources, community infrastructure) and identification of potential areas of conflict between these factors. By GIS analysis of the on-the-ground issues related to sustainability, the key parameters to be considered in decision-making were identified. This type of analysis is crucial not only for regulatory compliance, but also to ensure that the operation has obtained the ‘social licence’ to mine. To demonstrate this approach, a case study was conducted on a mountaintop coal mining operation in central Appalachia. High-conflict areas were found to be primarily concentrated near streams and residential developments through the use of GIS. The case study suggests that use of this approach could have allowed for better communications and planning. The controversies surrounding permitting at the site, including litigation, could have also been avoided. Adoption of such a process could assist in the transition to a new regulatory framework that promotes and is based on sustainable development principles.     

Use of mesh adaptivity in simulation of flow in packed beds – A case study

Computational modelling of transport phenomena in packed beds is of interest in a wide range of engineering applications. We simulate flow in packed beds with similar conditions to heap leaching processes. The governing equations are the conservation of mass and Darcy’s law. The primary variables are pressure and saturation. To discretize the governing equation, a control volume finite element method is used. The numerical scheme is implemented in Fluidity, a general purpose flow simulator that is equipped with anisotropic mesh adaptivity techniques to control local solution errors and increase computational accuracy. We demonstrate the application of the developed method to solve for incompressible flow of air and liquid within packed beds of mono-dispersed non-porous particles and compare our results with experimental measurements by Ilankoon and Neethling (2012). To study the effect of heterogeneity on flow regime in packed beds, an experiment is conducted on a packed bed consisting of two regions with different particle sizes. The non-uniform packed bed is also simulated numerically and our comparison shows very good agreement. Mesh adaptivity resolves saturation fronts in two regions of the column with high-resolution meshes and reduces the discretization error. This study serves as a validation for our numerical approach for prediction of hydrodynamics of heap leaching processes at large scales.     

How does the representation rate of features in a landscape affect visual preferences? A case study from a post-mining landscape

Previous studies of visual preferences for landscapes have been dominated by papers monitoring the presence of individual landscape features or their physical attributes. The objective of our study has been to evaluate how the representation rate of these features affects visual preferences for a landscape. We have studied a post-mining landscape that provides component physical features of a diverse and contrasting character. On the basis of an analysis of surface images, in combination with a questionnaire survey, our study has determined that the representation rate of most evaluated features affects the aesthetic value of the entire landscape scene. It has been established that certain groups of features in a post-mining landscape evoke various preferential tendencies among the public. The aesthetic value of a landscape decreased with increasing representation of surface quarries and urban structures, while increased representation of water and of rural structures resulted in an increasing aesthetic value. Woody vegetation and arable land were most preferred when they were represented non-dominantly, while mining infrastructure features were evaluated most negatively when they were non-dominant features of the landscape. Our findings thus point to the importance of the representation rate of features in the landscape as one of the key factors influencing the visual perception of the entire landscape scene.     

Natural vs. reclaimed forests – a case study of successional change,reclamation technique and phytodiversity

The present study accounts for the successional changes that took place in vegetation structure and composition over a time span of 25?years at reclaimed manganese mine site. Adjoining natural forest site was also studied for comparison of floral composition. The results corroborate the fact that as the age of succession increases, the reclaimed area gets invaded by higher successional species leading to an ecologically sound self-sustaining restructured system. This is an indication of the efficacy of the restoration programme. Plant species diversity of reclaimed site was 3.34, 2.81 and 2.12 for herbs, shrubs and trees, respectively, whereas in the adjoining natural forest site (Pench national park, reserved forest), it was 2.74, 1.88 and 2.13 for herbs, shrubs and trees, respectively. Phytodiversity assessments at both sites led to the conclusion that higher values of importance value index (IVI) were obtained for shrubs and herbs at the reclaimed site when compared to the adjoining natural forest site. However, higher values of tree diversity and IVI were associated with the natural forest site than the 25-year-old reclaimed site. This paper compares the composition of flora, species diversity and rapid succession of phytodiversity of a 25-year-old reclaimed site with nearby natural reserved forest site. Plant species richness of reclaimed site was 141, 38 and 18 for herbs, shrubs and trees, respectively, as compared to reserved forest site values of 55, 11 and 20 for herbs, shrubs and trees, respectively. Also, this paper evaluates the total and available nutrient, and microbial diversity in both the newly formed ecosystem and the adjacent natural forest site.     

Post-regrind selective depression of pyrite in pyritic copper–gold flotation using aeration and diethylenetriamine

The present study investigates the effect of aeration and diethylenetriamine (DETA) on the selective depression of pyrite in a porphyry copper–gold ore, after regrinding (at grind sizes, d₈₀ = 38 and 8 μm) with respect to Au recovery and grade using oxygen demand tests, flotation, QEMSCAN, X-ray spectroscopy (XPS) and EDTA extraction analysis. It was found that pyrite depression increases after aeration and with decreasing grind size. This was observed to be due to the markedly higher oxygen consumption rate of pyrite at the 8 μm (kla = 0.10 min⁻¹) than at the 38 μm grind size (kla = 0.02 min⁻¹). The addition of DETA improved pyrite depression (9% with aeration only versus 39% with aeration + DETA) at the 38 μm grind size. Gold and copper flotation recovery followed pyrite recovery for the two grind sizes using XD5002 in the presence of air and DETA.The surface analysis (XPS and EDTA extraction) revealed that the significant pyrite depression at the 8 μm grind size was due to increased amount of surface iron oxides, oxy-hydroxides (FeO/OH), sulphate species and increased liberation of mineral phases (QEMSCAN analysis), whilst the poorer pyrite depression at the 38 μm grind size was due to insufficient liberation of mineral phases and the persistence of activating Cu on the pyrite surface. The addition of DETA increased pyrite depression at the coarser grind size due to a significant reduction in Cu(I)S and increased Cu(II)O species, correlating with the flotation results of pyrite under this test condition. Two-stage copper and pyrite flotation, followed by Au cleaning after regrinding to 38 μm grind size, under high pH or aerated condition is proposed as the recommended route to optimise Au flotation.     

Optimisation of the secondary ball mill using an on-line ball and pulp load sensor – The Sensomag

The ball load and pulp load have a significant influence on the ball mill product size and production capacity. To improve the circuit performance at industrial scale these variables must be tweaked to levels where the plant can get grind and capacity benefits. In most grinding circuits the influence of these variables are not quantified because it is difficult to obtain precise measurements of the pulp load for an industrial scale mill and the conventional method of obtaining ball load measurements that involves crash stops is not attractive. A comprehensive investigation was performed on an industrial scale mill to quantify the effects of both ball and pulp load. A wide range of ball and pulp loads were tested and the findings are reported in this paper. The Sensomag, a sensor developed by Magotteaux, was used to obtain ball and pulp load measurements during the experimental work.     

A case study of optimising UG2 flotation performance part 2: Modelling improved PGM recovery and Cr2O3 rejection at Northam’s UG2 concentrator

In Part 1 of this 2 part series of papers the relationships between Platinum Group Metals (PGM), mass and water recovery and %Cr₂O₃ (as chromite) in concentrate were defined for laboratory, pilot and plant scales. The entrainment of chromite in final concentrate was shown to be related to its slow floating kinetics.In 2004, Northam Platinum mine embarked upon an upgrading program on its UG2 plant which included the installation of two column cells with external spargers as final cleaners. Optimisation resulted in a PGM recovery increase of 6% whilst %Cr₂O₃ in final concentrate was reduced from 4.0% to 2.2%.In this second paper, the Northam circuit is modelled before and after modification and installation of the final cleaner column cells. Simulation shows that the external sparger, driven by a dedicated recirculation pump, imparts energy into the system and provides the column with a PGM recovery capability equal to that of a mechanically driven cell. The deep froth bed of the columns enhances chromite rejection. As a result of this, greater operating flexibility has allowed PGM recovery at Northam to be increased by increasing mass pull whilst at the same time reducing %Cr₂O₃ in concentrate. Modelling shows that Northam’s reduction in %Cr₂O₃ can be fully described by reducing only the laboratory to plant scale-up factor for slow floating rate of chromite by 25%, whilst leaving the normal scale-up factors for fast floating fraction and rate unchanged. The improvement in PGM recovery was simulated using the same set of rougher and cleaner feed PGM kinetics and scale-up factors. This suggests that chromite entrainment is modelled via its slow floating rate and PGM recovery by true flotation is primarily modelled via its fast floating fraction and rate. Floatable gangue is modelled via its full set of fast and slow kinetics as its passage into final concentrate (and cleaner tailings circulating load) is a mix of true flotation and entrainment.This case study shows that a standard flotation circuit and one incorporating Northam’s column cell may be predicted from laboratory scale rate tests and flotation kinetics. The impact of Northam’s technology on other UG2 ores may be predicted by applying the change in chromite scale-up factors determined by simulation. The degree of improvement is dependant on the ore’s mineralogy and specifically its selectivity between PGM minerals, chromite and floatable gangue. This is illustrated by simulating the impact of the Northam column cell on Barrick’s Sedibelo UG2 ore.     

CFD modelling of bubble–particle collision rates and efficiencies in a flotation cell

Computational fluid dynamic (CFD) modelling of a Denver-type flotation cell has been performed. Bubble–particle collision rates in different parts of the cell have been calculated from the local turbulent velocities, and the size and number concentrations of bubbles and particles obtained from CFD modelling. The probability of collision due to streamline effect of fine particles moving around the bubble has also been estimated. The local attachment rate based on the collision rate and collision probability is then calculated and found to decrease as particle size decreases. This is consistent with the decrease in flotation recovery of fine particles as observed in flotation practice. The magnitudes of the collection rate constants obtained from CFD modelling indicate that transport rates of the bubble–particle aggregates to the froth layer may contribute quite significantly to the overall flotation rate in plant-scale equipment.     

The interaction of lignosulfonate dispersants and grinding media in copper–gold flotation from a high clay ore

This study was conducted to explain an inconsistent observation in the previous study that a lignosulfonate-based polymeric dispersant improved copper and gold flotation from high clay ores significantly in laboratory but clear improvement was not observed during the full-scale plant trial (Seaman et al., 2012). In this study the flotation of low and high clay ores after grinding with different grinding media was performed to understand the interaction of lignosulfonate dispersant and grinding conditions equipped with rheology measurements and EDTA (ethylene diamine-tetra acid) extraction. It was found that lignosulfonate dispersant mitigated the negative effect of clay minerals on copper and gold flotation when the amount of iron oxidation products originated from the grinding media was minimised. Mild steel media produced a great amount of iron oxidation products depressing copper and gold flotation and masking the role of the dispersant.     

Simulation of particle flows and breakage in crushers using DEM: Part 2 – Impact crushers

DEM simulation using a breakage model is applied to study a broad range of existing impact based crushers to better understand machine level operation and to demonstrate that such modelling is now possible. Predictions for a range of types of hammer mills, vertical and horizontal shaft impactors are presented. The breakage method used is based on fracture of particles using geometric rules for progeny generation and estimates of the stress applied to the particles as they flow through the crusher. From these simulations, estimates of power, product size, throughput rate and crusher wear are made.     

Comparative study of the explosion pressure characteristics of micro- and nano-sized coal dust and methane–coal dust mixtures in a pipe

Coal dust explosion accidents often cause substantial property damage and casualties and frequently involve nano-sized coal dust.In order to study the impact of nano-sized coal on coal dust and methane–coal dust explosions,a pipe test apparatus was used to analyze the explosion pressure characteristics of five types of micro-nano particle dusts(800 nm,1200 nm,45μm,60μm,and 75μm)at five concentrations(100 g/m3,250 g/m3,500 g/m3,750 g/m3,and 1000 g/m3).The explosion pressure characteristics were closely related to the coal dust particle size and concentration.The maximum explosion pressure,maximum rate of pressure rise,and deflagration index for nano-sized coal dust were larger than for its micro-sized counterpart,indicating that a nano-sized coal dust explosion is more dangerous.The highest deflagration index Kst for coal dust was 13.97 MPa/(m·s),indicating weak explosibility.When 7%methane was added to the air,the maximum deflagration index Kst for methane–coal dust was 42.62 MPa/(m·s),indicating very strong explosibility.This indicates that adding methane to the coal dust mixture substantially increased the hazard grade.     

Computational prediction of performance for a full scale Isamill: Part 1 – Media motion and energy utilisation in a dry mill

The Isamill is a horizontal stirred media mill used for fine and ultrafine grinding. Discrete Element Method (DEM) simulation has been used in this paper to study the behaviour of the media and the collisional environment in a full industrial scale Isamill. The DEM modelling has been able to provide a general understanding of the force balances that control the mill performance and the quantitative variation of key performance measures such as power draw and peak collision energy with changes in mill operating parameters and media properties. The key factor controlling the behaviour of the media and mill performance is the force balance on the media adjacent to the rotating discs of the impellor. Gravity and any applied axial pressure gradient play a critical role by providing a restoring force that presses the media bed against the discs which enables energy transfer from the discs to the adjacent media. Three important aspects of the media flow are identified with the holes in the discs playing a critical role in media transport and generating high speeds media jets in the upper half of the mill.     

Dependence of fluid flow on cleat aperture distribution and aperture–length scaling: a case study from Gondwana coal seams of Raniganj Formation,Eastern India

Dimensions and mutual relationships amongst fracture(cleat)parameters such as maximum aperture width,average aperture width,length,spacing etc.control the connectivity within a fracture network and the fluid flow in a coal seam as the matrix permeability,here,is negligible.In this paper,we document cleat size distributions and investigate length–aperture relationships from coals of Raniganj coalfield in Eastern India.This coalfield has a proven extractable reserve of six billion tons of coal and holds immense potential to be one of the largest coal bed methane fields serving India’s growing energy needs.Here,cleat length(L)correlates with corresponding maximum aperture width(Dmax)in a power-law function with an exponent of 0.84(DmaxαL0.84)instead of the commonly observed exponents of 1 or 0.5 applicable for other natural‘opening-mode’fractures.The conventional wisdom pertains that laminar fluid flow(Q)through an isolated,smooth-walled,parallel-plate fracture,embedded in an impermeable matrix,is directly proportional to the cube of its aperture width(b,equivalent to Davg;cubic law:Qαb3).This assumes a linear relationship between length and fracture aperture.However,the modified relationship between cleat length and average aperture width changes the cubic law applicable for Raniganj coal seam and now fluid flow correlates with aperture width in a power-law function with an exponent of 4.25(Qαb4.25)instead of 3(cube).Such simplifications will come handy for the modeling and estimation of fluid flow as it will reduce the effort of cleat length measurement which is anyway difficult and can be misleading due to the risk of undersampling.