Search Documents

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

Conditional simulation can be used to determine the probabilities of block iron-grades being above selected cut-offs. However, there are much faster alternatives that can produce acceptable results. In this paper blasthole data from a selected area of Joffre Member hosted ore of the Brockman Iron Formation at the Mt Whaleback orebody is used as the basis of a case study to assess alternatives to simulation. Iron grade, in per cent, is interpolated from a blasthole dataset into a block model using ordinary kriging. Samples are then removed from this blasthole dataset to produce data subsets. A complete distribution of iron grades for points or blocks is obtained utilising the data subsets and the following techniques: conditional simulation, median indicator kriging, the probability from conditional expectation and confidence intervals attached to the estimation of blocks in the framework of the discrete Gaussian model. The complete blasthole dataset is considered to represent the true iron grades at sample point support while the ordinary kriging estimate derived from these samples is considered to represent the true block grades. The distributions obtained from point estimates, block estimates and simulations are compared to each other and to the true estimates of iron grades using graphs and statistics. The conditional expectation technique reproduced the results obtained from conditional simulation extremely well and it is a viable alternative to simulation at point support. At block support the confidence intervals technique reproduced the results of the conditional simulation more closely than the median indicator kriging and in the data subsets tested the confidence intervals technique actually reproduced the true grade distribution more closely than the conditional simulation.

Jan 1, 2005

By R Berry

This paper looks at two ways to estimate the bulk mineralogy of the rocks for assay intervals. The aim is to find an efficient indicator of the most common minerals in the rock. Phase (modal) analysis has traditionally been done using visual methods such as point counting and image analysis. A modern version of this process is the X-ray point counting routine using the SEM-EDS based software. These methods are too slow and expensive for routine analysis of bulk sample mineralogy at the normal assay spacing.Two sources of data were considered that provide information that can be used to determine the mineral abundance in assay samples. The most widely applied method is (semi-) quantitative X-ray diffraction (QXRD). The QXRD method is most applicable to major minerals and has limited application to minerals at low abundance. The nominal detection limit is 0.5 per cent. Values below five per cent have large errors. A second, less common, method is calculation of mineralogy from chemical assay data. Conversion of chemical analyses to mineralogical analyses depends on the unique chemical composition of each mineral. Elements only found in one mineral are easily accounted for, but many compositions are ambiguous. Deciding on the actual mineralogy is not simple. Recalculation of mineral mode from chemical analyses is more accurate than QXRD when the correct minerals, and mineral compositions, are known.Where only a few QXRD analyses are available they can be used to setup a protocol for calculation of mineralogy from assay data. Linear programming works well in this environment. The best results are obtained when both H2O and CO2 are directly measured. Loss-on-ignition (LOI) should be included if these are not available.Where both QXRD and chemical analysis are available for all samples, the best results are obtained using the least squared method to merge the data sets assuming QXRD has much higher analytical errors than chemical assays. The combined method provides more robust results because the high abundance minerals are controlled by the QXRD measurements while the chemical assays improve the precision for low abundance minerals

Sep 5, 2011

By G A. Streltsova

In the paper the main concept of qualitative simulation is determined when computer-aided designing and forecasting of designs of robotised continuous-flow working face complexes for underground mining of hard ores. The parametric model æunderground ore mineÆ and its components û composite models are considered. Frame description of structure of working face complex and methods of organising of rule-oriented representation of knowledge when itÆs operating are given. The architecture of the developed prototype of expert system æworking face complexÆ and construction of a knowledge base for it are described when solving problems of multi-criteria optimisation of computational parameters of the working face complex under conditions of uncertainty. The guidelines for further development of knowledge base and expert system are given.

Jan 1, 2003

The subject of octanes and knock in petrol engines is of general interest, not only to the engineers and chemists in the petroleum and automotive industries, but also to the motorist who wishes to see some light through the haze of propaganda regarding "super" petrols. A mixture of petrol vapour and air will explode when a spark is passed through it if it contains more than one and a half per cent. or less than six per cent. of petrol vapour by volume. In a petrol engine the mixture has to be within the explosive range before it will operate. When the spark is passed it 'ignites the petrol vapour-air mixture which burns with rapid but finite progress of the flame front 'through the unburnt gases in much the same manner that a charge of dynamite is consumed. The turbulence of the gases in the cylinder under normal operating conditions more than doubles the velocity of the flame front and enables the combustion to be completed by the end of the working (expansion) stroke:KNOCKIf the cylinder head 'is poorly designed or is covered with a deposit of carbon, it is possible for "hot spots" to develop, and these can ignite the mixture before the spark is passed. This pre-ignition gives rise to an engine knock but is accompanied by a dull thud rather than the ringing knock associated with the true "knocking" found in a petrol engine and attributed to the detonation of the mixture. Until a few years ago it was thought that this detonation was due to a shock wave produced by the rapid combustion of the mixture, and sufficiently intense to cause ignition of...

Jan 1, 1951

By Whiting AN

Major changes in the world in recent years are changing the criteria for selection of future developments. Oil prices soared during the 1970's while commodity prices including metals and minerals fell. Many new independent nations, have been created and now control the resources in areas which were once part of great colonial Empires.

Jan 1, 1983

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 A V. Nguyen

Spherical calcium dioleate particles (~10 ¦m in diameter) were used as AFM (atomic force microscope) probes to measure interaction forces of the collector colloid with calcite and fluorite surfaces. The attractive AFM force between the calcium dioleate sphere and the fluorite surface is strong and has longer range than DLVO (Derjaguin-Landau-Verwey-Overbeek) prediction. The repulsive AFM force between the calcium dioleate sphere and the fluorite surface does not agree with the DLVO prediction. Consideration of non-DLVO forces, including the attractive hydrophobic force and the repulsive hydration force, was necessary to explain the experimental results. The non-DLVO interactions considered were justified by the different interfacial water structures at fluorite- and calcite-water interfaces as revealed by the numerical computation experiments using molecular dynamics simulation.

Jan 1, 2005

By C M. Standen

The SAG mill at Rasp Mine has undergone a number of liner design modifications since commissioning in 2012, making liner wear and failure difficult to predict from historical data. During 2017, several instances of liner failure or near failure were experienced resulting in unplanned downtime for emergency relines or to patch liners in the lead up to a planned shutdown. The strategy for relining mills is often to minimise the frequency of relines (and subsequent shutdown costs) by extending liner life, achieved through design modifications. However, it was observed that as the Rasp SAG mill liners wear, mill performance declined with power draw increasing and throughput declining. Opportunity and inefficiency costs start to erode the revenue per additional tonne milled for a given set of liners, and in turn the benefit of delaying relines. As such, optimising the frequency of relines becomes a trade-off between reline costs and opportunity/inefficiency costs in delaying relines. The aim of this paper is to demonstrate a frame work for analysing costs using a marginal economic analysis with the assumption that the Law of Diminishing Returns can be applied to the life of mill liners. The SAG mill operating variables can be modelled to determine their impact on costs and an optimal reline point can be defined where the total costs are minimised. This prevents unexpected failures due to accelerated wear by providing justification to reline earlier and thereby allowing efforts to be focused on sustaining milling performance.CITATION:Standen, C M, 2018. An economic approach to optimising mill reline frequency – a case study of SAG mill relines at Rasp mine, in Proceedings 14th AusIMM Mill Operators' Conference 2018, pp 197–206 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Aug 29, 2018

By Ito H, Noguchi F, Ueda Y

DTA curves of synthesis lead sulfide and commercial. lead concentrates during oxidation were measured to clarify the mechanism of ox- idation of lead sulfide. The oxidation reac tions could be divided to elementary reactions with production of basic lead sulfates such as 2PbO PbSO4. An existance of iron sulfide played an important role of the oxidation of lead concentrates. The oxidation rates of lead sulfide were measured in the wide range of oxygen potenti- als using IR type SO2 gas-meter. The oxidation rates under low Po 2 which corresponds to the direct oxidation smelting were much lower than those obtained under high Pot. The combustion temperatures of lead sulfide were also measured by hot thermocouple method. Those values were extremly low compar- ing the oxidation temperatures obtained by DTA.

Jan 1, 1987

The genesis of the Bulong Laterite Ni/Co deposit spans some 60 per cent of earth history: from development of the parent rocks in the Archaean to lateritic weathering in the Cainozoic. In a man's term, the discovery and evaluation history of the deposits has also been protracted and, at any given time, the perceived value of the deposits has depended upon the available technology and world markets.

Jan 1, 1996

The subject of government involvement in industry is, to say the least, a touchy one. For many it conjures up thoughts of socialisation or nationalisation. For others it is a desirable objective and is seen as a means of social reform by giving government a greater say in the control and distribution of human, financial and natural resources for the betterment of the community. Whatever your point of view, it is difficult to deny that government involvement exists to lesser or greater degrees in all industries. The mining industry is no exception and the degree of government involvement is probably greater than it is in other industries. I think it is important to remember when talking about government involvement in industry to recognise that there are different types of involvement and that all governments have a right, if they so desire, to become involved in whatever way they choose. The decision of a government to increase its direct or indirect involvement must, of course, be taken in the national interest. Governments, if we are still old fashioned enough to believe in ministerial and electoral responsibility, must make up their own minds about whether their actions are in the national interest and they must be prepared to accept the consequences of that action, good or bad. For instance, they must decide

Jan 1, 1975

By M S. Manono, K C. Corin

The optimisation and process developments surrounding the beneficiation of Cu-Ni-PGM ores is of fundamental importance when considering the continued rise in global demand for these precious metals. The selective sequential flotation of these ores is accompanied with various processing challenges such as the successful separation of the Cu-Ni-PGMs from non-sulfide naturally floatable gangue and pyrrhotite. Reagents such as sodium iso-butyl xanthate (SIBX) and sodium carboxymethyl cellulose (CMC) have proven to significantly improve the selectivity of the flotation process in their respective roles. The use of recycled water with abnormal inorganic electrolyte concentrations can modify the selectivity of these reagents, as ions may adsorb onto minerals such as pyrrhotite, thereby hindering or promoting the hydrophocity and floatability of pyrrhotite. Process water quality effects on the behaviour of pyrrhotite may have negative or positive influence on the recovery and grades of the wanted Cu-Ni-PGMs as valuable minerals associated with pyrrhotite such as pentlandite may be affected in like manner. Thus, the impact of concentrator process water quality on the behaviour of pyrrhotite needs to be determined. In this study synthetic process water at various ionic strengths was investigated against various CMC dosages to understand the behaviour of pyrrhotite during the flotation of a Cu-Ni-PGM Merensky ore in a batch flotation cell. This study showed that pyrrhotite grades increased with an increase in CMC dosage for fixed ionic strengths whereas the recoveries of pyrrhotite increased with an increase in ionic strength until 3SPW, then a trend of a decrease in recoveries was seen with 5SPW resulting in the lowest pyrrhotite recoveries. Furthermore, this study showed that moderate CMC dosages of 100 g/t resulted in the highest pyrrhotite recovery for all four ionic strengths of synthetic process water tested. It can be said that to maximise value recovery from Cu-Ni-PGM ores, understanding the behaviour of pyrrhotite against degrading water quality and reagent suites could be of paramount importance given the sensitivity of pyrrhotite to oxidising chemical conditions.

Aug 24, 2022

By R A. Pax

Operational grinding mills emit significant acoustic emissions which are a consequence of the dynamic environment prevailing inside them. From a technical perspective, the noise presents an opportunity to determine some details of the processes occurring inside the mills. The noise generated is not random and can be attributed to a range of events that occur inside these mills, consequently non-contact acoustic determinations of in-mill behaviour provide powerful insights into the operation of primary comminution tumbling mills. To enable a logically coherent picture to be formed of the processes inside a mill, arrays of microphones have been used. This approach allows a clearer picture of the in-mill behaviour of the charge to be accessed. In contrast to single microphone approaches, usually located near the expected charge toe position, an array of microphones allows much more useful information to be obtained from grinding mill noise.This paper will discuss and present data obtained from microphone arrays positioned at the toe and shoulder locations of a mill, and how this information can then be used for the optimisation of grinding mills in real-time.CITATION:Pax, R A, 2012. Determining mill operational efficiencies using non-contact acoustic emissions from microphone arrays, in Proceedings 11th AusIMM Mill OperatorsÆ Conference, pp 119-126 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Oct 29, 2012

By T Honeyands, L Jelenich

We are currently witnessing large fluctuations in the relative price of ironmaking raw materials brought on by changes in the dynamics of supply and demand, market conditions, environmental regulations and technical factors affecting the iron and steelmaking process. From a customer perspective, it is the behaviour of iron ores in downstream processing that give them their value; their impact on the sintering or pelletising process and subsequently blast furnace ironmaking. It is therefore important to consider this value when developing projects, making mine planning/cut-off grade decisions and in setting price differentials for differing quality.This paper describes the use of the Marx value-in-use model to quantify the downstream value of lump iron ores. The Marx model consists of heat and mass balance modules for sintering and pelletising and a rigorous two-stage heat and mass balance model of blast furnace ironmaking. Mass balance and cost models are applied for steelmaking, casting and rolling. The use of a heat and mass balance allows accurate comparison of the impact of raw material properties on blast furnace operation. The impact of minor elements, such as alumina, silica and phosphorus, and metallurgical properties on ironmaking is described, and examples are given for the calculation of a lump premium.CITATION:Honeyands, T and Jelenich, L, 2015. Lump ore properties and their impact on value-in-use in ironmaking, in Proceedings Iron Ore 2015, pp 359–362 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Jul 13, 2015

By M W. Chen, I Canbulat, S Saydam, C G. Zhang

The height of fracturing (HoF) measures the propagation distance of continuous fracturing from the mining horizon and is also termed the height of cracking (Hebblewhite, 2020; Khanal et al, 2019) or hydrologically the height of groundwater drainage (Tammetta, 2013). In longwall practice, HoF is an important predictive indicator for assessing the dewatering effect of panel extraction on the shallow aquifer and surface flows. The HoF determination is basically dependent on field measurement works, typically using surface extensometer, borehole televiewer imaging, piezometer monitoring, packer testing, borehole breakout logging, and isotopic tracer monitoring (Brown and Walsh, 2022; Corbett, 2022; Hebblewhite, 2020; Holla and Armstrong, 1986; Mills and O’Grady, 1998; Walsh et al, 2022a, 2022b). These methods or tools have been widely applied to observe rock fracturing and groundwater responses to longwalls at Australian coalmines, for example, Dendrobium Mine (Brown and Walsh, 2022; Walsh et al, 2022b), Appin Mine (Walsh et al, 2022a), and Springvale Mine (Corbett, 2022). In Chinese mines, borehole televiewer imaging and packer testing are more frequently used due to convenient instrumentation, quick testing, and not requiring long-term maintenance of the instrumentation boreholes. The HoF determination also involves further raw data interpretations, for which specialist software for data extraction and quantitative criteria for bounding the fracture transition and termination are necessary. By combining the measured fracture distribution and longwall parameters regarding panel geometry and geological settings, several HoF models have been developed for conceptualising the understanding of strata depressurisation and providing mine engineers fast calculation tools for HoF prediction. The Mackie model considered the height of fracturing as the upper bound of fractured zone and revealed that within the fractured zone there are highly connected fractures and accordingly porosity enhancement and groundwater depressurisation, where the fracture network can be less transmissive from the mining horizon up (Hebblewhite, 2020). The fractured zone shares similar fracturing regime but different heights in varying models: HoF is 30 to 58 times of the mining height (t) based on Australian, Soviet and UK experience (Kendorski, 1993, 2006), and 21 to 33 t according to the field measurements in NSW Central Coast, Australia (Galvin, 2016). Some Australian longwall mines adopt Ditton model (Ditton and Merrick, 2014) (Equations 1.1 and 1.2 in Table 1) sourcing from NSW coalfield experience for HoF prediction, while Liu model (Liu, 1981) (Equations 2.1 to 2.4 in Table 1) developed from 27 coalmines in China is frequently used by Chinese coalmines.

Nov 29, 2022

Average annual growth in consumption of primary energy in Australia over the past 10 years has been 5·5 per cent compared to a world average of 4·1 per cent.Over the next 10 years growth rate is expected to increase to 6 per cent per annum due largely to the rapid development in the mineral processing industry.By the end of the 19705 natural gas is expected to supply more than 9 per cent of Australia's primary energy market. Coal will supply 40 per cent and petroleum products 47 per cent. Nuclear energy will make its appearance but will make only a small contribution in the period.Australia has adequate reserves of natural gas and coal to supply its needs in the 1970s and beyond. However the newly developed reserves of crude oil, while supplying 60-70 per cent of requirements in the initialyears, will need regular and sizable additions if they are to keep pace with the expanding consumption. For Australia to become self sufficient in the period 1970-1990 it is estimated that new discoveries will need to approximate ten times the volume of recoverable reserves found in Australia to date.

Jan 1, 1971

By Patridge A. C

A pilot scale laboratory investigation of manual slurry sampling methods representative of those used in the minerals industry is described. Samples were taken from slurries made up from silica particles of various sizes and in various concentrations from a stirred tank and from the discharge lip of a rectangular launder. Sampling methods from the tank included immersion flasks and several configurations of a suction probe. The launder was sampled using a conventional narrow-mouth ladle and also with a 'mouth organ' stream splitter. Samples were analysed for overall solids concentration and size distribution and showed different but always significant and often large biases relative to the known overall composition of the slurry, the nature and extent of bias being dependent on pulp concentration, particle size, sampling method and sampling location. Results clearly show the very great difficulty of obtaining representative samples by simple sampling techniques, although the data suggest that errors can be reduced by using an appropriately designed and operated suction probe. Further work will be carried out with the objective of establishing guidelines for industrial sampling practice.

Jan 1, 1995

By W Valery, A Jankovic

A growing demand within the minerals industry for sustainable development and the introduction of carbon taxes have resulted in a significant focus on improving the energy efficiency of comminution circuits. However, in this drive for improved energy efficiency the role of classification has been overlooked. Inefficient classification and large recirculating loads reduce circuit capacity, result in over-grinding and higher specific energy consumption.It is well understood that classification increases mill capacity; therefore, closed ball mill-cyclone circuits have become an industry standard. However, due to limitations in cyclone classification efficiency, the optimum recirculating load is limited to about 250 per cent. A simple relationship quantifying the effect of classification efficiency and recirculating load on grinding circuit capacity is presented in this paper. The possibility of improving the capacity of ball mill circuits by improving classification efficiency is discussed, considering the current practical limitations of classification equipment. Simulation tools were used to evaluate circuits with cyclones and with high frequency screens to identify opportunities to improve energy efficiency. The investigation also identified potential benefits to flotation recovery due to a reduction in over-grinding.CITATION:Jankovic, A and Valery, W, 2012. The impact of classification on the energy efficiency of grinding circuits - The hidden opportunity, in Proceedings 11th AusIMM Mill OperatorsÆ Conference, pp 65-70 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Oct 29, 2012

By B Kilgariff

Flotation has been in use at Broken Hill for nearly 100 years, and the process has gone through many different optimisation studies over that time. This paper reviews the current flotation practice at the Perilya Broken Hill Mine, and details one of the recent optimisation studies conducted using mathematical modelling and simulation tools. The optimisation study consisted of surveys of the lead and zinc circuits, mapping of the floatability components of the ore and developing models of the lead and zinc circuits. The models were then used to conduct simulations to simplify the circuit. The study allowed plant personnel to identify optimum operating conditions for their current flotation cells, as well as possible improvements in circuit configuration and cell capacity.

Jan 1, 2005