Search Documents

By N Gruber

In the continuous casting of ultralow carbon (ULC) steels, free carbon is used to control the melting behaviour of mould powders. If the carbon is not completely removed during melting, it is enriched at the top of the slag pool. Liquid steel may come into contact with this layer because of the turbulence of the molten metal, resulting in its recarburisation, which negatively affects the desired product quality. Thus, a reduction in carbon input is desirable. For this purpose, SiC and/or Si3N4 with and without antioxidants were selected as melt-control additives to replace carbon in the mould powders. Thermodynamic calculations were performed to quantify their effect on the melting behaviour based on the chemical composition of a flux already applied to ULC steels. To experimentally assess the liquefaction behaviour, laboratory mould powders were prepared and annealed in steel crucibles closed with a lid. Crucibles were inserted into a furnace that was already preheated to selected temperatures between 900–1200°C for 10 mins and quenched to room temperature. Subsequently, the samples were mineralogically investigated. The results confirmed those obtained from the thermodynamic calculations. Si3N4, and SiC in particular, are suitable raw materials for delaying the solid-solid reactions of raw material components during melting. Owing to their stability at high temperatures, the necessary SiO2 content to form a liquid phase is not available, resulting in lower amounts of the liquid phase. The addition of antioxidants to delay the oxidation of SiC further reduces this positive effect. Attempts to decrease the SiC content without negatively affecting the melting behaviour resulted in a reduction in the CO2 emission by at least 27 g CO2/kg of mould powder when compared to the carbon-containing standard mould powder. These investigations revealed differences in the melting behaviours of granules and loose powders, which are related to their respective production processes.

Jun 19, 2024

By M Nagarajan, S Esen

This paper presents three case studies from three surface coalmines. The first case study is from a coalmine where a dragline operates in 60 m wide and 1 km long strips. Typical overburden height ranges from 30 m to 50 m. The base case blast was modelled using an advanced blasting model (Distinct Motion Code – DMC) to establish the baseline using the blasting parameters in use before optimisation. DMC was used to model the alternative blasting scenarios to obtain the following key objectives:the post-blast muck pile profile was required to have an optimum height of ~28 moptimise cast percentageachieve a well-fragmented muck pile for the draglinethe muck pile profile should be appropriate for the dragline entry into the strip.Achieving these objectives would optimise the dragline performance through reducing rehandle and improving advance along the strip. DMC modelling resulted in the following outcomes:the required muck pile profile could be achievedthe cast percentage increased from 21.1 to 25.1 – a four per cent increaserehandle volumes were reduced.Establishing a good quality assurance / quality control (QA/QC) process was essential to delivering a blast that could achieve these outcomes. The blast was fired successfully and the dragline completed the strip approximately two weeks ahead of the time taken for the previous strip.The second case study is from a coalmine where D11 dozers are used to move the blasted material to the spoil. An optimum profile was required for the dozers for this low-cost mining operation. To achieve this, an electronic blasting system (uni tronicTM 600) was introduced to optimise the muck pile profiles being delivered. The new muck pile profiles showed improved cast percentage from 10.6 per cent to 17.8 per cent. The changes in the muck pile centre of mass displacement improved dozer efficiency as a result of pushing material shorter distances to its final position. Dozer productivity (bcm/h) statistically did not change but coal was being exposed at a faster rate.The third case study summarises the results from a site where dozers are used to minimise the total mining cost. Traditionally, excavators had been used on-site with conventional blast designs. After firing the first cast shot, the blasted material was found reasonably well-fragmented. Although dozing productivity hasn’t been quantified on-site, operators were providing positive feedback (easier push) due to the reasonably good muck pile profile suiting to the dozers. This site also realised significant savings with the dozing as compared to their truck/shovel operation.CITATION:Esen, S and Nagarajan, M, 2015. Muck pile shaping for draglines and dozers at surface coalmines, in Proceedings 11th International Symposium on Rock Fragmentation by Blasting, pp 409–416 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Aug 24, 2015

By M R. Davidson, M P. Schwarz

A model for column bioleaching is investigated to identify and understand aspects of bacteria in bioleaching applications, which have implications in heap bioleaching operations. The model is used to simulate scenarios that would otherwise be time consuming to perform experimentally. This study uses a model of bacterial transport and attachment/detachment to ore particles, with a bioleaching model for the depletion of a copper-sulfide, also accounting for liquid and gas flow and gas/liquid oxygen mass transfer. The model includes aspects such as oxygen and ferrous ion consumption, coupled with leaching of a copper-sulfide via the shrinking core model. The model is used to investigate how the rate of leaching is affected by the bacterial concentration in the columns, by the bacterial regeneration of the leaching oxidant ferric ions. The model is also used to assess the impact of oxygen limitation and inoculation method on the copper leaching. Some comparison of the model with experimental data will be shown.

Jan 1, 2004

By J Nortier, J Dray

There is no single factor that determines the appropriate capacity for a mining project. It is driven by the size of the resource, the capacity of the market, the availability of capital, the availability of power, the availability of water, and a myriad of other considerations. There is also no single solution to the question of how big the project should be. However, to develop a project, the study team needs to select the configuration options that it regards as the best for the project. To be able to make this selection, the study team should consider the range of available options and do so at a level of detail that gives it confidence in the conclusion but does not lead to multiple detailed design exercises. Having considered the range of options, and with a clear definition of the project owner’s objectives and constraints, the work to be carried into feasibility studies can be focused appropriately. This paper discusses the process of completing strategic mine planning to identify project configuration options to assist in the selection of the processing plant throughput rate (Capacity Study) for a large greenfield open pit mine. The processes described in this paper allow for the basis of design to be set for subsequent studies that meets the objectives at a risk profile acceptable to the owner.

Jun 22, 2022

By J R. Manuel, M J. Peterson

Selected drill core samples from a Ferrobamba-limestone-hosted, magnetite-haematite skarn deposit, located in the Apurimac department of the Andahuaylas-Yauri skarn/porphyry belt of south-eastern Peru, were mineralogically and texturally characterised to gain a preliminary indication of the deposit's potential as an iron ore resource. Optical microscopy, X-ray diffraction (XRD) and scanning electron microscopy/electron probe microanalysis (SEM/EPMA) techniques were utilised to determine the characteristics of magnetite, haematite and/or goethite dominated ore types, including variations in the relative proportions and textural relationships of the ore and gangue minerals. The deportment of elements deleterious to processing performance (eg copper, sulfur and potassium) was determined and the beneficiation potential of the samples was briefly assessed. In drill core sections, the ores are grey to reddish in colour, massive and 'metallic' to dark brown and earthy in character, moderately hard and porous (macro and microporous), with variable gangue infill. Ores often appear weathered, typified by ochreous goethite staining and common minor goethite (except in dense core fragments), occasionally exhibiting a more friable texture with coarse magnetite/martite grains. Silicates and lesser carbonate gangue are visible as blue-green to beige to white scaly and/or shaly surface coatings, and finely particulate infill, often goethite stained. Sulfide-rich ore types also occur in part of the deposit, but are restricted to specific drill holes. Magnetite grain size is relatively coarse, compared with magnetite from typical banded iron formations, whereas discrete silicate gangue is mainly bladed/fibrous and/or layered and micaceous in form. XRD analysis indicates the presence of vermiculite, serpentine, stilpnomelane, chlorite, actinolite and clays. There is a spectrum of ore types representing the gradation from 'fresh' (magnetite-dominated) to oxidised and weathered (martite/goethite-dominated). Martite-dominated samples are generally higher in grade, due to leaching and removal of silicate gangue, resulting in higher macroporosity. Where gangue has been leached out, partial replacement by vitreous goethite and/or hydrohaematite is typical, often preserving the fibrous/layered characteristics of the original silicates.

Jan 1, 2009

By D Kentwell

For deposits where a full assay suite of elements/oxides is available and where several variables are critical to the evaluation of the deposit, it is desirable for the estimated block values of all elements/oxides to sum to 100 per cent of the weight per cent. This is difficult to achieve in practice and the estimation methods used need to be designed to minimise bias of the key elements/oxides by ensuring that block estimates are consistent with the individual assays and that the estimated totals are as close as possible to 100 per cent.This paper presents a case study of two domains within a magnetite hematite deposit, one with a high level of drilling classified as Measured and one with a lower density of drilling classified as Indicated. Several estimation methods were tested to evaluate their ability to maintain consistency of the proportions of each element/oxide and sum to 100 per cent. The following methods were evaluated: ordinary kriging, co-kriging, inverse distance and min/max autocorrelation factors (MAF) all as implemented in the Geovariances Isatis software.The comparisons showed that the MAF method produced the most robust estimates in terms of block totals and that grades and tonnages at the reporting cut-off were not significantly different between methods. At an individual block scale all methods produce artefacts that are less than desirable but can only be remedied by additional drilling.CITATION:Kentwell, D, 2015. Comparison of min/max autocorrelation factors, kriging, co-kriging and inverse distance block totals – a magnetite hematite case study, in Proceedings Iron Ore 2015, pp 465–474 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Jul 13, 2015

By A D. Campbell

Sublevel caving (SLC) is a mining method adopted in a wide range of mining depths and ground conditions with various layouts adopted to suit orebody geometry. Sublevel caving offers more flexibility and versatility compared to other cave mining methods and is a low cost alternative to open stoping. SLC also has the advantage that stopes are not required to remain stable during production, and no backfill is required. This flexibility combined with increasing pressure on mines to target deeper and lower grade mineralisation has led to increased uptake of the method. The implementation of SLC mining methods across a wide range of mining depths and rock mass conditions has led to ground control practices which vary globally from mine to mine. To date, there has been limited documentation of ground control practices in SLC mining and few benchmarks available in the literature. This paper describes the main findings from a global benchmark of sublevel cave mines conducted over three years as part of the Caving 2040 research consortium. The purpose of the project was to document and compare operational practices, technical aspects and hazard management techniques across SLC mines around the world. The review included 21 mines spread over four continents and 17 different mining companies. This paper provides an overview of geotechnical conditions, challenges and ground control practices at each of the benchmarked SLC mines. Mine design and sequencing practices adopted for ground control purposes and hazard mitigation are provided with case studies. The impact of depth and rock mass conditions on mining productivity and development rates is also described.

Nov 29, 2022

By S Döpp

Commencing with early experiences in iron ore beneficiation made at Beeshoek Mine, the paper follows through the process of extensive mineralogical and metallurgical investigations carried out during feasibility study. The paper will discuss the design process and logic followed over the last years, the challenges and decisions the design team faced in order to draw the conclusions that ultimately resulted in today’s value add approach at Assmang’s Khumani Operations.As integral part of the benefi ciation concept, a focus is set on the jigs, explaining the plant concepts, planned expansions and their contribution to the overall value add approach at the mine.The case study is intended to describe the outcomes of the design versus the actual achieved and the further value add projects to be implemented. As for future prospect, the incorporation of Wet High Intensity Magnetic Separation (WHIMS) demonstrates Assmang Khumani’s pursuit to maximise productivity and overall recovery.DMS was evaluated as an alternative to jigging within a comparative test work study, but in conclusion the option had to be excluded due to its inability to achieve the high density cut point required to achieve the sales specifi cations. In contrast, the investigation on producing an additional middlings fraction on the lumpy ore jigs prove to be an adequate option for further improvements of overall mass yield. With further liberation, a saleable product will be achieved by secondary beneficiation on Fines Jigs and WHIMS.Concluding, jigs are considered the most viable and cost effective benefi ciation process for off-grade ore +0.8 - 32 mm at Beeshoek as well as at Khumani, while WHIMS is the process under consideration for super fines recovery.

Jul 11, 2011

The homologous series of saturated fatty acids investigated in this study generally show a steady gradation of properties with chain length. The solution pKa values were assumed constant at 4.9, however reported pKa values could be lower or higher. Hematite flotation response and surface pressure for aqueous solutions of these fatty acids were measured as a function of pH. For the shorter chain fatty acids, the undissociated acid was shown to be the predominant surface active species with flotation, supplemented by the physical adsorption of acid anions as the chain length increased. The reduced solubility of the longer fatty acid carbon chains made these reagents ineffective as flotation agents. The maximum flotation response of hematite using tetradecanoic acid corresponded with the precipitation edge of the reagent. The results show that the values of surface pressure of the surfactants are not always a good indication of their effectiveness as flotation collectors.

Jan 1, 2005

By M Stephens, T A. Sutra

Currently, there are several open pit mines with production rates of about 200 000 tpd of ore that are reaching the end of their economic lives. In most cases, the remaining orebody extends to depths greater than 300 m below the pit limit, leaving hundreds of millions of tons of reserves that can only be recovered by means of an underground mining method. One such method used for mass production is block caving. This paper presents a ventilation system for a four-zone block cave mine operation developed on a footprint 1000 m wide and 1200 m long. The system is designed for a mine with a nominal production rate of 120 000 tpd of ore using diesel powered units on the production and reduction levels. It discusses the air quantity requirements for the worst case scenario, which is anticipated to occur five years after the mine has reached its full production capacity. Finally, it summarises the results of computer modelling exercises that were carried out to determine an economic ventilation infrastructure that includes shafts, drifts, main fans, and ventilation control devices.

Jan 1, 2005

By C Lower

An unprecedented, quantitative, deposit-scale understanding of major clay mineral distribution throughout the Yeelirrie uranium deposit has been achieved by adapting the existing HyLogger™/HyChips™ analytical technique to measure resource assay pulps. Spectral analysis via HyLogger™ and HyChips™ systems has previously been used as a small-scale geometallurgical research tool, but never before has it been quantitatively applied to an entire resource sample population for the purpose of developing stand alone block models for major clay minerals. Economic uranium mineralisation within the thin but extensive Yeelirrie?palaeochannel?primarily exists within the transition zone between a groundwater-derived calcrete and the underlying clay-quartz regolith. Early geometallurgical investigations identified that certain clay minerals could be deleterious to the efficiency of the uranium recovery process. Accurate calculation of clay mineralogy from resource assays proved too difficult due to chemical similarities between co-existing kaolinite, smectite and magnesium carbonate minerals. A relatively low-cost, efficient and accurate method of quantitative clay mineral analysis was required to measure over 70 000 resource assay pulps. After initial testing and comparison with QXRD, the HyChips™ 3-1 spectral logging system was implemented as a primary method of clay mineral analysis for the Yeelirrie uranium deposit. The resultant spatial data set allowed the interpretation and geostatistical analysis of geometallurgical domains across the deposit. Clay block models were generated and used in geometallurgical resource calculations to benefit the mine planning process.

Sep 5, 2011

By J Ferreira

To speed up multivariate geostatistical simulation it is common to transform the set of attributes into spatially uncorrelated factors that can be simulated independently. One decorrelation method is the method of minimum/maximum autocorrelation factors (MAF), where a two structure linear model of coregionalisation is assumed. The transformation decorrelates the theoretical model exactly, but the actual data are only approximately decorrelated. The MAF transformation can therefore be seen as a special case of a non-orthogonal approximate diagonaliser of a set of symmetric matrices. A more general approach for approximate joint diagonalisation (AJD) has been developed in the context of blind source separation. For these AJD algorithms no assumptions are made beyond symmetry of the individual matrices and so they can be applied to a family of experimental semivariogram matrices. In their application there are no restrictions on the number of structures in the linear model of coregionalisation (LMC) thus removing one of the conditions placed on the subsequent modelling of the spatial structure of the factors. In this paper the general background for the U-WEDGE AJD method is presented. It is shown that the U-WEDGE method achieves better spatial decorrelation than the MAF method. The factors derived from U-WEDGE are simulated using conditional turning bands simulation and then backtransformed to attribute space.

Sep 26, 2011

By H J. Ross, R K. Beatson, T J, W R. Fright

Traditional method of explicitly defining three-dimensional (3D) ore-waste and geological boundaries relies heavily on a time-consuming process of manual digitisation. This method of modelling can be best described as surface modelling, as complex surface geometry is built up by digitising points that lie on the surface. With the advent of fast 3D interpolation methods, however, construction of geological surfaces using volume functions is now a practical alternative to explicit modelling of surfaces. Unlike explicit modelling, surfaces contained in volume functions are not explicitly defined or digitised. Instead the existence of surfaces in the volume function is implicit, thus the process of modelling surfaces from volume function is called æimplicit modellingÆ. Based on recent advances in fast scattered data interpolation methods, implicit modelling first defines a continuous three-dimensional function that describes the grade or rock distribution. This volumetric function is interrogated for a grade value, or a geological surface, thus allowing the extraction of the 3D object to be automated and eliminating the need to manually digitise surfaces. Since the function is continuous throughout space and does not depend on a mesh or grid for its definition, the extracted geological or grade wireframes can be constructed at any desired resolution in the specific volume of interest. The volume modelling method can work on scattered drillhole data of any data density, including processing combined information from dense grade control data as well as sparse resource drilling. It offers distinct advantages over surface modelling, including: being able to wireframe geological objects that may take days to digitise in tens of minutes; substantial improvement in modelling accuracy; and the ability to generate conditional models rapidly allowing mining risks inherent in geological modelling to be examined.

Jan 1, 2003

It is well documented that drill and blast outcomes affect the downstream mining processes of loading, hauling, crushing and grinding but it can also significantly impact on mine profitability via dilution, recovery percentage, lump generation (iron ore), fines generation (coal, iron ore), coal loss and final wall stability. Drill and blast must therefore be optimised to improve the overall profitability via positively impacting on productivity and saleable product generation. It is however, difficult to achieve this because of the complexity of the number of parameters that must be integrated and managed across the entire mine value chain, ie the concept of æMine to MillÆ. This paper highlights, via a simple case study, the need to have a mineÆs drill and blast system both in control (so that it is predictable to a high degree) and with minimal variability (so the design actually makes it into the field). This type of management control is essential prior to optimisation work being undertaken. The æTeam ApproachÆ is discussed to illustrate the methodology of achieving drill and blast optimisation. A case study is presented to illustrate the potential benefits of this approach which in this mine represented a 27 per cent reduction in powder factor with the downstream functions achieving higher productivity with the same or reduced maintenance costs. It is essential for Australian mineÆs to implement the appropriate management strategies and experienced personnel placement to ensure that their mineÆs drill and blast system is correctly managed to facilitate the optimisation of profit. Without this, the Australian mining sector will become increasingly uncompetitive.

Jan 1, 1999

By Chunbao Z, Zhenchuan Z

As known to the international mining sectors, Hunan province, called as ' Nonferrous Native Area ', is the main base of nonferrous metals Pro- duction in China. There are over 194 nonferrous mines in this province, most of them distribute over the basin of Xiangjiang River. The state of management of wastewater in Hunan's nonferrous mines is outlined.The experiences and economic benefit in the management and treatment of wa- stewater in Bofang and Mayang copper mines, and at Shizhuyuan polymeta- llic mine, are mainly introduced. It concludes with measures and recommendations for future management and treatment of wastewater in Hunan's nonferrous mines.

Jan 1, 1988

By Song Yongxue

During the design and operation stages of a block caving project, a large amount of research and development work has been undertaken in the Tongkuangyu Copper Mine. This paper presents the research achievements from these studies, including a rock mass quality evaluation method, cavability prediction, fragmentation evaluation, caving mechanisms, a computerised production scheduling model and the optimised draw control system. The main features of the block caving system in the Tongkuangyu Copper Mine are also discussed.

Jan 1, 2000

By E C. F Hamman

Sunrise Dam Gold Mine is a large open pit mine 730 km NE of Perth, Western Australia. Contract mining commenced in the open pit in September 1996 in lake clays and saprolites with the first fresh rock mined late 1999, and the underground starting the first decline in October 2003. By May 2007, over 150 million cubic metres of material had been mined and the pit was 360 m deep with 80 m remaining before reaching the planned final depth in early 2008. A final cut-back in the northwest will commence in late 2007 with open pit mining in the Mega Pit expected to conclude in late 2010. The challenges of developing the pit can be divided into three phases:Soft materials û mining was in multiple staged pits and cut-backs in lake clays, saprolites, weathered material. The varying characteristics had a significant impact on mining methods and slope stability. Structural û mining was gradually exposing more fresh rock and the drive was to understand how major structures and lithological contacts impact on local and large-scale stability. This had a significant influence on mine design and scheduling. Slope angles û achieving the final slope angle is critical. This is done through application of high standards of final wall establishment and proactive geotechnical hazard identification and management. This includes dealing with the rapid rate of descent, interaction with underground excavations and managing rock fall risk. The paper discusses each of these phases, highlighting the geotechnical controls implemented to date, looking at their success and impact on mining.

Jan 1, 2007

By E Jak, M Shevchenko, D Shishin

One of the key aspects to develop sustainable metallurgical production is to ensure that the predictive power of thermodynamic tools is brought up to a new level of accuracy and reliability. Exploring new polymetallic processes, integrating primary and recycled materials, means utilising the uncharted areas within the Cu-Pb-Zn-Fe-Ca-Al-Mg-Si-O (major) – Cr-Na (slagging) – As-Sn-Sb- Bi-Ag-Au-Ni-Co (minor) slag-solids-metal-matte-speiss-sulfate system. This requires extensive integrated experimental and thermodynamic modelling study, which is underway at PyroSearch at the University of Queensland (UQ).

Aug 21, 2024

By D J. Clark, R F. Wright, M D. Goodz

The A1 Gold Mine is a Devonian-age narrow vein high-grade gold deposit located 115 km east of Melbourne. The mine operated almost continuously from 1861 to 1992 producing 442 334 ounces of gold (13.8 t) from 530 436 tonnes from primary gold ore tonnes processed at an average recovered grade of 25.9 g/t Au. The A1 dyke was attributed with a yield exceeding 620 000 ounces (19.3 t) of gold (including gold in alluvial and tailings deposits), and was renowned as one of VictoriaÆs premier mines paying considerable dividends. The orebody was principally in a diorite dyke bulge 150 m long, 45 m wide and 700 m deep and was mined via two vertical shafts. The mine site was closed in 1992 and the site was rehabilitated in 1999. In 2008 - 2009, a diamond drilling program identified various extensions to the host dyke lithology and to gold-bearing quartz vein filled fault zones over a 700 m strike length. The outcome of this study was the establishment of a mine geology corridor four times longer in strike length than the original 1970s model. This opens A1 dyke hosted deposit style to exploration outside the historical dyke bulge and provides exploration opportunity for all similar deposits. Construction of the first 3D geological model has developed new insights on the host rock diorite dyke system and identified various extensions to both host stratigraphy and to gold-bearing mineralised structures. Coarse visible gold in quartz vein mineralisation (>30 g/t Au) and medium-grade stockwork mineralisation (8 - 15 g/t Au) has been intersected within 100 m of the surface and up to 300 m along strike of the nearest historical underground workings. At depth it appears that the mineralisation is continuous with the main dyke body with consideration of minor rotation in the strike orientation and limited reverse thrust faulting showing maximum displacement of 10 - 20 m. Along the A1 mine corridor, between two and five parallel dykes are observed at different drill sections. The gold mineralisation is hosted mainly within the dyke rocks, but where the sedimentary rocks are intensely silicified or are xenolith rafts, the gold mineralisation continues across these sedimentary rock blocks. There has been inference to the role of silicified sedimentary rocks as a second ore host and a bounding material for inflections in fault and vein orientations. In the central 150 m long A1 mine dyke bulge, the extremities are demarcated by four to five parallel dyke fingers (referred to as apophyses) going both north and south. These apophyses are clearly observed both vertically and horizontally between levels No 4 and No 9. The dyke fingers entrap sedimentary rock xenoliths and collectively form the A1 dyke æcorridorÆ, which is quite uniform in overall width from the easternmost contact (footwall) to the westernmost contact (hanging wall). Dyke fingers were assumed to coalesce at depth into a single dyke, but recent drilling shows this feature to repeat in several positions with depth and along strike. Key observations are the relationship between gold grade and fracture density, alteration, sulfidation, host rock morphology û relation of contact shape and fault zones. These relationships allow for the classification of prospectivity indicators and geology as a guide to potential extensions of mineralised zones. Mining geology is contributing to the whole-of-mine life planning and development where decline path, stockpile and infrastructure positions are controlled by geology.

Jan 1, 2009

By J J. Cilliers, K Hadler

Air is fundamental to flotation; not only is it required to collect the valuable particles, but it also controls the froth stability and consequently the separation performance for a given feed. Froth stability can be described and quantified in numerous ways, which are subtly different. Air recovery is one such measure of froth stability which accounts for both the stability of the bubbles and the mobility of the froth. The benefit of using air recovery as a measure of froth stability is that it exhibits a peak as air flow rate is increased, and this corresponds to the air rate at which the optimal separation performance is obtained. As such, it provides a single target value for optimising air flowrate for a given set of feed conditions. Use of air recovery, however, has one potential drawback: accurate measurement of the inlet air flow rate to a flotation cell is required. This can be problematic for operations, particularly where calibration of air flow rate is carried out infrequently or where there is no air flow rate measurement, such as for many self-aerated flotation cells. To this end, we present an alternative method for determining the peak in air recovery (PAR) air rate.We introduce a simplified method for determining the PAR air rate using trends in froth velocity and overflowing froth height, both of which are readily measured using existing froth image analysis systems. We will show, using industrial data, the application of this simple technique to find the optimal air flowrate for flotation cells. We will also discuss other alternative techniques for air rate optimisation that have the potential to improve separation performance using existing technology.CITATION:Hadler, K and Cilliers, J J, 2018. How to find the optimum air rate for flotation cells, in Proceedings 14th AusIMM Mill Operators' Conference 2018, pp 377–384 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Aug 29, 2018