Search Documents

By A. Mainza, A. Mabentsela

A finite element model of a 5 m radius DC arc ilmenite furnace in idling mode was used to test the notion that the slag solidifies when it comes in to contact with colder pig iron, thus constituting the initial step in the formation of solid slag at the slag-pig iron interface. It was found that a slag that is 150°C hotter than the pig iron does not solidify at the interface. The 150°C temperature difference between the slag and pig iron is a result of solid slag at the slag-pig iron interface, not the other way around as suggested in the literature. Calculations show that the thickness of the frozen slag at the slag-pig iron interface is 1.7 cm for the furnace used. It is proposed that slag solidification begins with the slow co-current flow of molten slag and pig iron in the outer parts of the furnace. This provides enough time for molten slag to interact with molten pig iron without solidifying. As the reduction products form due to reduction of the slag by carbon in the pig iron, the slag solidifies. This approach negates the need for the slag to solidify by merely coming into contact with an inherently colder pig iron. Making use of a low thermal contact conductance between the slag and pig iron was found to be sufficient to numerically capture the presence of solid slag at the slag-pig iron interface and to preserve the 150C difference between the slag and pig iron phase.

Mar 30, 2023

By F. Wang, P. Rosario, A. Vazirizadeh

Advances in data science, machine learning, and artificial intelligence are transforming mining and mineral processing, and making them more algorithm intensive. The paradigm is shifting from one of detection and control to one of prediction and optimisation. To build up reliable prediction and optimisation models these steps should be passed: collect, aggregate, cleanse, and process large amounts of complex, structured, and unstructured data as a first stage. Next, the right tools must be used to extract meaningful features from the operational data for the build-up of a prediction model. A grinding circuit is one of the challenging areas in terms of determining its overall efficiency and the prediction of product quality. In a processing plant, some of the KPIs are typically unmeasurable or infrequently measured for real-time application, e.g. ore grindability and grind size. However, these variables are required to calculate the grinding circuit efficiency and to model the operation performance. In this paper, advanced analytic tools have been used to extract one selected KPI (cyclone overall P80 or final grind size) and its relationship with measured variables for an industrial dataset. The results illustrate that the performance of the grinding circuit can be predicted for real-time monitoring by means of measured data. Keywords: Grinding circuit, data-driven, machine learning, process modelling, hybrid modelling

Jan 1, 2020

By J. Dvorácek

Similar to the case of any other business venture, the activities of mining industries are profit-oriented. Nothing will stop successful businesses in their quest for new localities to be prospected for minerals of their choice, because the useful economic life of any mine is limited. Before starting the process of land acquisition, it is important for any mining company to be able to forecast the profitability of the venture. The focus of this paper has been on the prospects of success, concerning the case of the opencast extraction of aggregates for construction. A set of 70 successful and loss-making quarries was subjected to investigation. An international company owns the quarries, which are located in seven European countries. The method of discriminate analysis provided the criteria for assessing future economic performance. The same criteria served the purpose of classifying other quarries belonging to the company.

Jan 1, 2012

By N. Sutcliffe

Refractories and slags play a vital role in the production of stainless steel. They are in intimate contact with the steel at temperature, and chemical reactions between the three can and do occur. These interactions can be beneficial, in the case of active tundish and ladle slags, and negative in the case of refractory erosion and slag entrapment. This paper presents the results of both plant trials and laboratory investigations of steel/refractory and slag/refractory interactions. The interaction mechanisms observed are described leading to selection criteria for new refractory products. Keywords: slags, steels, refractories, reaction mechanisms.

Jan 1, 2004

By A. P. Squelch

The question of what stress condition exists in the hangingwall of tabular stopes remains topical and largely unanswered. Limited field data exist to quantify the situation and provide calibration for investigative computer modelling exercises. Work conducted in1989 and 1990 under a Chamber of Mines Research Organization (now the CSIR: Division of Mining Technology) programme of research into the behaviour and benefits of backfill may assist with this situation. The original work was motivated by the premise that the improved hangingwall stability in backfilled panels (identified by rockfall accident analyses) is due to a change in the hangingwall stress regime. Current, largely theoretical, studies into mechanisms of stope hangingwall behaviour and support interaction could benefit from considering these data. Keywords: Backfill, horizontal stress, stope hangingwall, hangingwall stability, support interaction.

Jan 1, 2002

By J. Gummer, M. Williams, S. J. Hawksworth

A common problem encountered in British coal mines is bearing failure on conveyor system idler rollers, which on occasions has led to fires. The UK Mines Inspectorate initiated a study of such failures to assess the potential for ignition of explosive atmospheres. As part of this study HSL have developed a system to measure the temperatures generated within the bearings during controlled failure. Temperatures are measured using both infrared thermal imaging and contact thermometry. The work to date has shown that temperatures well in excess of the auto-ignition temperature of methane can occur.

Jan 1, 2003

By M. K. Sichone

Anglo Platinum?s Waterval smelting complex based in South Africa operates the Anglo Converting Process (ACP) that produces SO2 offgas. This SO2 offgas is treated in a double-contact, double-absorption acid plant allowing for conversion efficiencies higher than 99.90 %. The catalytic converter used in the acid plant to convert the SO2 to SO3 is a 4-bed reactor loaded with a combination of cesium-promoted and conventional vanadium pentoxide catalysts. In June 2008 there was a shutdown, in which one of the main activities was to screen the catalyst. Before this shutdown, the operator would control the converter pass temperatures manually, via the SCADA, especially during plant start-ups after which some of the automatic temperature controllers were used. After the shutdown, the control philosophy was revisited and all temperature controllers were put on auto mode during plant start up. Furthermore the set points of the pass inlet temperatures were reduced to maximize conversion in the first two passes. The result has been a reduction of SO2 emissions, from a daily average of 50 ? 70 ppm to as low as 20 ppm for similar sulphur dioxide and oxygen concentrations. Further work has been done to maintain high bed temperatures in the third and fourth passes at low SO2 strengths by splitting the gas stream between pass 1 and 2 thus increasing the SO2 content in the gas to pass 2. This increases the temperature of the gas exiting the second pass, which results in higher temperatures in passes 3 and 4. Comparison of the converter performance prior to and after the changes is made. The benefits of partially bypassing the first pass to increase the SO2 content in the gas to the second pass are shown. Conclusions and scope for further work have been drawn from the comparisons.

Jan 1, 2009

Significant advances have been made in the design of centrifugal slurry pumps that are used for pumping high-density viscous and paste slurries. This paper presents the results of a series of centrifugal pump performance tests that were conducted using an FLSmidth 6 x 4 Krebs millMAX centrifugal pump. The pump was fitted with a dry mechanical seal to prevent dilution of the slurry from gland service water, as the theology of the slurry is sensitive to small changes in slurry density. Tests were done using paste kaolin slurry with Bingham yield stress values between 578-1 120 Pa and plastic viscosities between 728 and 1491 mPa.s. The viscous properties of the paste during each pump test were measured using a rotational viscometer. Pump head and efficiency de-rating were measured and compared to the correlations of Walker and Goulas (1984). The test data shows the pump's best-efficiency point (BEP) changes and shifts to the left of the performance curve. The pump efficiency decreases as the slurry yield stress increases. The head de-rating data shows more head de-rating at the low flow rates compared to the high flow rates for the high viscosity slurry, which is in line with the findings of Walker and Goulas (1984). This paper demonstrates that it is possible to pump very viscous, high yield stress slurries with limited head and efficiency de-rating using centrifugal pumps, provided that positive suction conditions are maintained at all times.

Jan 1, 2012

By C. Thiart

Kriging is an interpolation technique for predicting unobserved responses at target locations from observed responses at specified locations. Kriging predictors are best linear unbiased predictors (BLUPs) and the precision of the BLUP is assessed by the mean square prediction error (MSPE), commonly known as the kriging variance. Both the BLUP and the MSPE depend on the covariance function describing the spatial correlation between locations and on specific parameters. The parameters are usually treated as known, whereas in practice they invariably have to be estimated and the empirical BLUP (that is, the EBLUP) so obtained. The empirical or estimated mean square prediction error (EMSPE), or the so called ?plug-in? kriging variance estimator, underestimates the true kriging variance of the EBLUP, at least in general. In this paper five estimators for the kriging variance of the EBLUP are considered and compared by means of a simulation study in which a Gaussian distribution for the responses, an exponential structure for the covariance function, and three levels of spatial correlation ? weak, moderate, and strong ? are adopted. The Prasad-Rao estimator obtained using restricted or residual maximum likelihood (REML) is recommended for moderate and strong spatial correlation and the Kacker-Harville estimator for weak correlation in the random fields.

Jan 1, 2014

By M. Pawlik, N. Ghaffari Touran

The interactions between apatite and hematite were studied through zeta potential and stability experiments, for this paper. The zeta potential measurements were conducted in both background electrolyte and mineral supernatants. The difference in electrokinetic properties of minerals in the presence and absence of supernatant was explained in terms of the dissolution of minerals and the subsequent readsorption of dissolved species at the mineral-water interface. The stability measurements in individual and mixed mineral systems were also performed to identify the extent of interaction between apatite and hematite upon their direct contact. The results from stability measurements in individual and mixed mineral systems indicate that there is an interaction between apatite and hematite. Based on the results, the implications for the apatite-hematite separation process were discussed. Keywords: Apatite, hematite, dissolved ions, stability, zeta potential

Jan 1, 2020

By F. Terblanche, M. Holton, A. Kruger

The paper presents an optimal mine planning system designed to simultaneously optimize all the underground mines of large South African gold and platinum mining corporations. The system, code named PlanIT-OPTIM, is a multi-stage solution involving the use of adetailed 3D CAD-based mine planning system (CADSmine™V4). The mine design and scheduling rules defined in CADSmine™ are responsible for simulating mining sequencing, rock mechanics, and ventilation, etcetera. Feasible production ranges generated by CADSmine™ and various high-level mathematical models are used as inputs to a mathematical optimization engine. The optimization engine produces optimal production profiles (i.e. key business indicators such as tonnages, mineral and areas extracted per time period) by maximizing net present value. Other objectives such as profit continuity, production continuity and maintaining available reserves can also be specified as optimization criteria. The final step in the PlanIT-OPTIM process is to convert the optimal production profiles into practical and attainable business plans. This is achieved through the CADSmine™ Detail Optimization Engine that modifies the original CADSmine™ schedule to match the optimal production profiles. With current methods it is an impossible task to find the best business plan for a mining corporation in any reasonable period of time should conditions change (e.g. gold prices, exchangerates, grade estimates, etcetera). The fast turn around time achieved through PlanIT-OPTIM makes it possible to do several ‘what-if’ scenarios in a reasonably short time. PlanIT-OPTIM has already been used in the gold mining industry to find optimal business plans for a mine.

Jan 1, 2003

By Y. Shiraishi, Y. Sakurai

A new method, the gliding cylinder method, was proposed as a wide range viscometer for the measurements of molten salts, slags, and glasses. The method consists of two coaxial cylinders, cylindrical bob and a vessel containing a liquid sample. There is a narrow gap between bob and vessel. When the vessel moves against the bob, the generated force on the bob consists of three types of resistance: (1) Stokes’s resistance on the bob, (2) a capillary flow resistance though an annular gap between vessel and bob, and (3) lubricating resistance in a gap. The last two can be calculated on a theoretical basis but the first cannot be estimated by a simple calculation since a wall effect did not calculate easily. However, it can be assumed that the generated force depends on both the viscosity of a liquid and velocity of the vessel. Thus, the relation between viscosity and a force at desired velocity of the vessel was measured using several standard reference materials including glycerol aqueous solutions and standard reference oils. The results depended on the magnitude of the gap between the bob and the vessel. For the wider gap, it was shown that the viscosity is a linear function of a gradient of force with respect to a vessel velocity. In contrast, if the gap is small, the viscosity depends on the gradient (of force with respect to velocity). This implies that the non-linear phenomena would occur in Stokes resistance. Though detailed analysis is in a further investigation, a calibration method can be applied to determine the viscosity by this method, because good correlation was obtained among these quantities. Viscosity of molten salts, such as NaNO3and ZnCl2was measured by this method as high temperature demonstration. Keywords: viscometer, viscosity, sphere pull-up method, fused salts, molten slags, glass melts

Jan 1, 2004

The spatial distribution of copper grades at the Mansa Mina project is highly variable due to the presence of high grade copper/silver/arsenic breccia bodies. Copper grades within the breccia bodies tend to average about 4.5% copper while grades in the surrounding stockwork average about 1.0%. The breccia is known to occur in the vicinity of significant controlling structures, however, the outline and therefore the thickness of the bodies are far from regular. Due to the high grade of these bodies a significant proportion of the value of the project is derived from the breccia tonnages. The irregularity in the outlines/thickness translates into an uncertainty in the breccia tonnage and therefore in the value of the project. Acknowledging the economic importance of the breccia bodies and the risk associated with the breccia tonnage uncertainty, significant additional drilling has been proposed to both understand the spatial distribution of the breccias and reduce project risk. The costs associated with this additional drilling are not inconsequential; thus, a reasonable question is ‘what value is derived from this additional drilling’? To obtain an answer to this important question, a simulation approach is applied. First the deposit is simulated using the existing drilling and the uncertainty in the resource is defined. Next the proposed drilling program is simulated and the simulated data are added to the existing data. Using this expanded data set a second simulation of the deposit is created and once again the uncertainty in the resource is defined. Using this approach, the reduction in the uncertainty in the resource can be determined. To convert this reduction in uncertainty (or reduction in risk) in the stated resource into a dollar value, a method of defining the increase in the project rate of return necessary to account for risk is presented. Using the two different rates of return, the difference in project NPV and hence the value of additional drilling is determined.

Jan 1, 2003

By Y. Rajpal

The Hedley and Grant pillar strength formula is widely used within the South African mining industry to calculate pillar strengths in bord and pillar mining layouts, and is currently being used at Bathopele Mine. Bathopele is an operation of Anglo American Platinum near Rustenburg, mining the UG2 reef at depths ranging from 40 m to 300 m below surface. The formula includes a parameter K, which is a downrated uniaxial compressive strength value of the rock, used to calculate the strength of a pillar. At Bathopele Mine, for the shallower areas Less than 150 m the value was 33 MPa, but this was increased to 44 MPa due to the fact that rock mass conditions improved and there were no incidence of failure in the shallower areas. Underground observations showed pillars to be stable with no signs of failure, and it was questioned whether the design was not still too conservative, resulting in sub-optimal extraction ratios and consequently a loss of revenue to the mine. A trial mining section was started at the mine where the pillars were reduced from 6 m x 6 m to 5 m x 5 m and the pillar behaviour was monitored for the onset of failure. Numerical modelling was conducted using the TEXAN code to back-analyse pillar stresses in order to find a more realistic value for K. Results from the trial section indicated that the pillar design was in fact too conservative, and the value for K was adjusted to 54 MPa for future designs. This paper details the findings from underground monitoring and numerical modelling results, and includes details for further calibration of the K value in order to further optimize the extraction of the orebody.

Jan 1, 2012

By V. Sheikhzadeh, S. M. H. Haddad, M. Saghaeian, N. Esmaeili, R. Hejazi

One of the most significant features of iron ore concentrate (pelletising feed) is the specific surface area (SSA) that affects the achievement of desirable pellet properties. There are various methods to determine this parameters and currently, the most common is the Blaine air permeability method. However, due to the simplified assumptions including spherical and identical particles and 50% bed porosity, the Blain test does not provide accurate results. In fact, these two suppositions are far from reality. In this research, comparison of SSA measurement ways are investigated, containing laser diffraction testing and Blaine air permeability, on samples with different SSA, obtained from Ball mill and high pressure grinding rolls (HPGR). The output data showed that the calculated amounts of SSA by means of laser analysis were far greater than any other one. Furthermore, in the same results of Blaine air permeability tests on samples comminuted by different grinding methods, it was observed that the SSA figures from the laser diffraction method for HPGR products were almost 800 units greater than Ball mill ones. Keywords: Pelletising feed, specific surface area, Blaine air permeability, laser diffraction testing

Jan 1, 2020

By P. van der Heever

The genesis of Simrac?s technology transfer practices and future initiatives are discussed in this report. Simrac produces a relatively large volume of research-based health and safety outputs that must be communicated to its tripartite stakeholders and to other interested and affected parties. The wide spectrum and complexity of the transferables, together with the diversity of end-user needs and skills of the recipient population, necessitate customizing of technology transfer practices. In an attempt to optimize this process, a variety of practices are employed. Information is transferred by means of publication and dissemination of a variety of reports (paper, website and CD).Knowledge transfer is directed to health and safety practitioners by means of workshops, seminars, best practice booklets, handbooks and textbooks. Promotion of recently completed projects is directed at targeted recipients by means of project launches. Only isolated cases of piloting and implementation of technologies have been undertaken due to lack of resources and past organizational focus. It is anticipated that Simrac?s new organizational structure and management system, together with end-user involvement by industry champions, from project inception to implementation, will enhance the effectiveness of the technology transfer process.

Jan 1, 2002

Electrochemical measurements are presented on electro-deoxidation of silica in molten calcium chloride at 900°C, with porous silica pellets as starting material. Measurements during pre-electrolysis (to remove water from the melt) showed electrochemical noise, which is consistent with the evolution of hydrogen. Throughout, the dominant anodic reaction appears to be CO2formation, based on the observed potential at the anode, and measured anode wear. Because of the presence of dissolved CaO in the melt, some of the CO2 candissolve in the melt; the resulting calcium carbonate formation was confirmed by electrolyte analysis after reduction runs. Reduction of dissolved CO2 at the cathode is likely to contribute to the background current during the reduction process. For the electrode and cell configuration used here, most of the over potential arises from polarization of the anode, and ohmic potential drop in the electrolyte.

Jan 1, 2006

By P. K. Hofmeyr

Base Metal (Cu, Ni, Mo, Zn, Co etc.) analysis traditionally has been very slow in conventional analytical laboratories which can lead to additional costs being incurred in waiting time of drill rigs or mining excessively off-reef. Recent developments in the automation of analytical laboratories now allow for very rapid turn-around times for base metal analytical data which can help reduce such costs as well as reduce the unit costs of the analyses themselves. 2.0 CONVENTIONAL ANALYTICAL LABORATORY PRACTICE Conventional analytical laboratories are necessarily are very manual. Sample preparation is slow and laborious, samples are moved from stage to stage manually and, typically, all samples are processed in batches so a sample which is the first to be, say, milled has to wait for the complete milling of the batch before it advances to the next stage eg. mixed with wax prior to pelletising for XRF analysis. Inevitably this results in long sample data turn-around times for laboratories handling large numbers of samples. Manual sample preparation can only be accelerated by increasing staff which increases costs and can lead to additional errors being introduced. Cleaning of sample preparation equipment, eg. crushers and mills, between samples is essential to avoid cross-contamination between samples and this is slow and labour-intensive. In manual sample preparation the correct sample numbers staying with the samples is totally dependent on the efficiency of laboratory operators. It is extremely difficult to detect when errors have occurred due to the switching of sample numbers during the sample preparation stages. Running duplicates and QC samples at best is only partially successful in detecting when sample number switching errors have occurred. Errors of this nature and cross-contamination problems due to incomplete cleaning of sample preparation equipment tend to occur when laboratories are working at pressure due to high sample numbers and when supervisors are not present during 24/7 laboratory operations which is often required on plant process control laboratories. These problems are common to all conventional analytical laboratories and their magnitude is dependent on the level of control and efficiency of QC systems practised by individual laboratories. The disadvantages of conventional manual analytical laboratories can be summarized as follows:- 1. Labour-intensive which is costly especially when multiple shifts have to be used during 24/7 operations.

Jan 1, 2005

By J. Martinovic

The large proportion of gangue minerals present in platinum bearing ores in South Africa provides challenges to the flotation and downstream processes. In order to take effective measures to reduce their recovery in flotation, an understanding of the mechanisms and factors influencing their behaviour is needed. Typical gangue minerals are pyroxene, feldspar, quartz, chromite and talc. Of these, talc is a naturally hydrophobic mineral and although it is present in relatively small quantities, it has a disproportionate effect as it has a strong froth stabilizing effect, increasing the recovery of other gangue minerals by entrainment. Talc recovery is effectively reduced with the addition of polymeric depressants. Although the other gangue minerals are assumed to be hydrophilic and their recovery by entrainment is expected to follow the water recovery, there is some evidence that they may be activated by the typical reagent suites used. Reagents such as copper sulphate, added as an activator, and collectors are used to selectively induce hydrophobicity on the sulphides. It has been shown that some of these gangue minerals can be activated. This research work investigates whether inadvertent copper activation is mineral specific and evaluates the mechanisms and factors affecting the reversal of activation with the addition of the depressants using electrokinetic measurements, adsorption and microflotation tests. Results are presented that show that the selected minerals do respond differently and that the reversal of activation is dependent on mineral type, depressant type, dosage and ionic strength of the electrolyte.

Jan 1, 2005

By M. Reuter, U. Boin

The target of the current metallurgical industry is to recycle and utilize all their by-products, so as to close the sustainable production loop. Slags are the important wastes and by-products of metallurgical industry, which have been treated, recycled and used worldwide. The present paper summarizes the current status of utilization of the various slags from ferrous and non-ferrous metal production, as well as waste incineration, and recycling of salt fluxes in secondary metal production. In addition, the environmental issues of the metallurgical slags are addressed. The metallurgical oxide slags have stone-like properties and, thus, their major applications are in civil engineering field. The slags should be recycled, modified and processed in a proper way, by taking the environmental impact into consideration. With the treatment of salt slags for melting aluminium scraps, as an example, the recycling and elution properties of the salt slags are discussed. Our research indicates that the viscosity of the salt flux is increased with addition of non-metallic components, that a certain percentage of fines remain in suspension, which determines the viscosity, affecting the settling of heavier materials. High slag viscosity will lead to more fine aluminium metal entrapped in the salt slag (may also be seen as a high temperature slurry), and thus increase the load of salt slag recycling. Keywords: recycling, environment, metallurgy, slags, fluxes, salts.

Jan 1, 2004