Search Documents

By Niklas Willemsen

Introduction to SVCs ? SVC = Static var Compensator ? Dynamic step-less control of reactive power without switching of passive components: ? Voltage control/stabilization ? Harmonics- and flicker mitigation ? Dip mitigation ? Main components: ? TCR (Thyristor Controlled Reactor) ? Tuned harmonic filters ? Control- and cooling systems ? Typical applications: ? Mine hoists ? Large mining complexes ? EAF, SAF, rolling mills, wind/solar farms etc..

Aug 1, 2013

By Fred W. Brackebusch

"The use of paste technology for surface tailings disposal may offer economic, technical, safety, and environmental advantages for both underground and surface mining operations. Paste is a non-Newtonian fluid having some shear strength during transport and placement. The strength is augmented after placement by consolidation, water loss, and by addition of chemicals. Economical dewatering systems are necessary to the success of surface paste tailings disposal. In an underground mine, the paste backfill plant can produce paste for surface disposal about 50% of the time because only about 1/2 of the tailings are needed for backfill. The distribution of paste to the surface stack involves expensive pumping equipment and thus a nearby location of the stack is desirable. The viscosity of paste can often be manipulated by modifying pH. Placement of the paste in a designed stack configuration depends on topographic and climatic factors as well as paste properties.INTRODUCTIONThe use of paste technology for mine backfill is becoming accepted for underground mines mainly because of ground control and economic reasons. Paste technology is just starting to be applied for tailings disposal with only a few hardrock metal mines worldwide using the technology; however, it has been used extensively for red mud disposal in the aluminum industry. The disposal of dry tailings is very attractive from environmental and safety points of view because of limited water which might potentially escape and pollute river systems. Dry stacking systems, however, are expensive because of filtration and truck transportation costs. Paste can be produced from most tailings streams by special types of thickeners. Furthermore, it can be transported in pipelines, therefore paste technology may be more economic than dry stacking but still have some of the major advantages. Understanding of paste technology has improved through experience with operating underground backfill systems in recent years and because of testwork and research programs by mining companies, consultants and equipment vendors. Thus we may be on the verge of taking a new look at tailings management using paste technology."

Jan 1, 2002

By D. Komljenovic, J. Paraszczak

"This paper presents the use of queuing theory in truck fleet sizing in open pit mining operations. In the first part, a methodology is developed by which an optimal ratio between the size of shovels and trucks may be determined. The second part gives the fundamentals of the queuing theory and its application in analyzing shovel and trucks. The stochastic character of involved times (loading, truck cycle, etc.) is taken into consideration. Using the above theory, the optimal truck fleet size in actual conditions may be estimated. Minimal operating cost of the working system (shovelstrucks) represents the basic requirement of the analyzed process. The methodology was validated on Caterpillar front-shovel – truck combinations. IntroductionLes camions représentent probablement les équipements les plus utilisés pour le transport dans les mines. Le coût de transport peut parfois atteindre 50 % du coût d’exploitation et 25 % du coût total (Collins, 1996; Kennedy, 1990). Il est critique pour une opération minière de le minimiser. C’est un sujet qui a été traité relativement souvent dans la littérature spécialisée. Les premières applications informatiques dans l’industrie minière ont été développées pour la détermination du temps de cycle des camions.Au fil des ans, la capacité des camions a augmenté et, aujourd’hui, les plus gros manufacturiers de camions (Caterpillar, Orenstein & Koppel, Euclid, Terex, Unit Rig, Komatsu, Liebherr, Belaz) fabriquent des camions ayant une capacité de 200 t et plus dont quelques-uns d’une capacité supérieure à 320 t. Pour les plus gros camions, le coût d’opération peut même dépasser 240 $US/heure dans des conditions modérées (Caterpillar, 1999). Les mines modernes de grande taille peuvent posséder plusieurs dizaines de camions dans leur flotte."

Jan 1, 2004

By D. Komljenovic

This paper presents the use of queuing theory in truck fleet sizing in open pit mining operations. In the first part, a methodology is developed by which an optimal ratio between the size of shovels and trucks may be determined. The second part gives the fundamentals of the queuing theory and its application in analyzing shovel and trucks. The stochastic character of involved times (loading, truck cycle, etc.) is taken into consideration. Using the above theory, the optimal truck fleet size in actual conditions may be estimated. Minimal operating cost of the working system (shovels-trucks) represents the basic requirement of the analyzed process. The methodology was validated on Caterpillar front-shovel ? truck combinations.

Sep 1, 2004

By Eric Konigsmann

"The Pogo project is located on the Goodpaster River, 38 miles north of Delta Junction in east central Alaska. Sumitomo Metal Mining and Teck Corporation are planning to develop a 2,500 ton per day underground mine and mill at Pogo. A preliminary resource of 5.5 million ounces of gold has been identified in two flat lying lenses. An underground adit has been developed to facilitate underground drilling of the deposit and to provide samples for bench and pilot scale metallurgical testwork.Flowsheet development included the evaluation of several processing options. The final flowsheet selected includes SAGhall mill grinding, gravity concentration with intensive cyanidation and direct electrowinning, sulfide flotation with ultrafine regrinding, and cyanide leaching and CIP gold recovery from the flotation concentrates. Flotation tailings will be filtered for placement in a dry stack storage facility or disposed of as paste backfill underground. All of the CIP tailings will be disposed of underground as paste backfill. A pilot plant was carried out to evaluate the proposed flowsheet described above and to confirm the gravity recovery predictions.Great care was taken in the development of the flowsheet and site layout concepts to mitigate the environmental impact of the proposed operation. An overview of the site water management concepts is presented and the impact of water issues on the process flowsheet is discussed."

Jan 1, 2002

By Brad McBain

Agenda ? Accurassay Laboratories and Automation ? History and experience ? Features and benefits of automation in the lab ? Tata Steel Minerals Canada (TSMC) Iron Ore Laboratory ? State-of-the art systems and design ? The integration of fusion systems and XRF units ? Meeting deliverables: quality, reliability, safety

Aug 1, 2013

By S. Cossio, J. L. Noyola, R. Espinosa

"Minera Penmont’s La Herradura is a joint venture between Industrias Peñoles, S.A. de C.V. and Newmont Mining Corporation. The mine is located in the northern Mexican state of Sonora. La Herradura consists of an open pit mine with a two-stage crushing plant and heap leaching facility. Gold is recovered from the pregnant solution using the Merrill-Crowe process; production began in July 1998. La Herradura produces about 130 000 oz gold per year and its treatment capacity is approximately 22 000 t/d. Ore mineralization occurs in a shear zone with quartz and native gold. Gold distribution is quite heterogeneous due to coarse visible gold, which in some cases ranges from 100 to 600 microns.Sampling and assaying protocols were established during exploration and prior to start-up of operations, based on nomographs developed from early heterogeneity tests. However, as the coarse gold problem became noticeable, it was necessary to develop new sampling protocols to have an acceptable relative standard deviation to control the fundamental error within ± 15%.An experimental program was developed and conducted on a selected ore block consisting of 195 blastholes; regular ore grade control procedures were applied to both sets of data to evaluate the difference in results between the protocol established prior to start-up of operations (existing) and the new protocol (leaching). In the existing protocol, sub-sampling is done through riffle splitters, and coning and quartering. Fire assaying was used to estimate the gold content on a 50 g sample. In the leaching protocol, a 12 000 g sample of bottle-rolled cyanide was leached, and the leach residue was assayed on a 50 g sample using fire assaying, and a weighted average total gold content was calculated. Currently, a new alternative method is being evaluated, where sub-sampling is done using rotary splitters, and fire assay on a 50 g sample; in this rotary sampling protocol and for comparison purposes, gold assays were also repeated using the leaching protocol.Test results were compared in three groups: < 0.3 g/t (below cut-off), 0.3 to 0.8 g/t, and >0.8 g/t. Overall, results show that the protocol established prior to start-up of operations underestimates the gold assays. Detailed technical results, alternative options, and their implications are discussed in the paper."

Jan 1, 2004

By M. Schmidt, P. Nord, M. Larinkari

Cathode stripping machine technology has been developed significantly both in electrorefining (ER) and electrowinning (EW) applications in the last ten years. The main driver has been robotic cathode handling, which has enabled increased flexibility to meet different layout and material handling requirements for both ER and EW processes. Additionally, cathode stripping stations have been continuously developed to optimize speed and to reduce mechanical stress of permanent cathodes. 'taco' type cathodes have been proven as beneficial approach for high capacity operations (particularly in Cu ER), where over 600 cathodes per hour continuous throughput has been achieved. However, the latest developments also apply to split type cathode stripping. Besides material handling, the modern cathode stripping machine has also an important role to provide information for production and quality monitoring. This paper describes the latest developments from the field of copper cathode stripping based on several different projects implemented in the recent years. INTRODUCTION After introduction of the permanent cathode technology in the late 1970s, it became the prevailing technology in copper refineries over the traditional starting sheet process by the end of 1980s. The main driver was less labour-intensive production due to fewer processing steps required along with more automated electrode handling equipment (cathode stripping machine, anode preparation machine, anode scrap washing machine and tankhouse cranes). Since then the design of the cathode stripping machine has followed the development of permanent cathode technology. The main development steps so far can be summarized as follows: Full Deposit Stripping Machine (FDSM) based on waxed permanent cathodes or cathodes with bottom edge strips resulting in split cathodes (copper sheets separated, see Figure 1) Wax less cathode stripping based on V-groove on bottom of the edge strip resulting in either ‘taco’ cathodes (enveloped copper sheets, see Figure 1) or split cathodes Full Deposit Robotic Stripping Machines (FDRSM) using 1–5 robots for permanent cathode handling and cathode stacking.

Jan 1, 2019

By Dean Lindsay, Inna Dymov, Mariam Melashvili, Jim Smolik, Chris Fleming

"This paper presents results from the proposed cyanide detoxification circuit for Seabridge Gold’s KSM project. The proposed flowsheet uses a two-stage CCD (counter-current washing/decantation) circuit to transfer the flotation products CIL(carbon-in-leach) barren solution to a SART process (Sulphidization, Acidification, Recycling of precipitate, and Thickening of precipitate) to recover copper and free cyanide from the copper cyanide complex. The SART barren solution is then treated by an AVR process (Acidification, Volatilization of HCN gas, and Reneutralization) to concentrate the cyanide that was liberated in SART for recycling to leach. The washed CIL barren pulp (the CCD underflow) is treated with the SO2 /Air process to oxidize residual cyanide.The goal of this complex flowsheet is to recover and recycle as much cyanide as possible and produce a final effluent containing less than 0.5 mg/L copper and CNWAD at the lowest cost. The SO2/air process is capable of achieving low residual CNWAD (<1 mg/L) in the final product, but the persistence of aqueous copper above the target level has been detected in the treated solution in some instances. The expected high concentrations of copper and cyanide in the KSM CIL barren ( >200 mg/L each species) and the difficulties this might create for the SO2 /air process was a key factor in the decision to consider including the washing, SART and AVR steps in the project flowsheet.The target concentrations of CNWAD and Cu were readily achieved after aging the SO2 /air treated pulp, as would occur in a tailing or polishing pond. This suggests no additional treatment will be required to sustain low levels of these species in the final plant effluent."

Jan 1, 2015

By C. A. Rowland Jr

"I. INTRODUCTIONFred C. Bond's ""Third Theory of Comminution"" (1) was published in 1952. The mathematical statement of the Third Theory is the well-known Bond Formula:where W - Kw H per short tonWl - Work IndexP - Product size in microns which 80% passesF .. Product size in microns which 80% passesWork Index, as defined by Fred Bond, "" ... is the comminution parameter which expresses the resistance of the material to crushing and grinding. Numerically, the Work Index is the KwHrs per short ton required to reduce the material from theoretically infinite feed size to 80% passing 100 microns."" (2)Work indices for grinding can be obtained tram Bond grindability tests and can be calculated from plant grinding mill operating data when power consumed per ton of ore ground, the feed size analysis and the product size analysis are known.Work indices obtained tram Bond grindability tests, having been obtained from standard and controlled laboratory procedures, are reproducible and for the sample tested are accurate. The degree of accuracy relative to plant mill operation depends upon how representative the sample is of the total mill feed.The results of the Bond closed circuit rod and ball mill grindability tests (3) are given in grams of product sized material produced per revolution of the test mill. In developing a method to evaluate plant operating data and results from his grindability test, Fred Bond developed the ""Third Theory of Comminution."" In the process formulae were developed for converting grams per revolutionobtained from grindability tests to work index.The standard feed sizes for the Bond grindability tests are:Rod Mill Test ore crushed to minus 1ll inchBall Mill Test ore crushed to minus 6 mesh (Tyler)The formulae for converting grams per revolution to work index incorporate the 80% passing sizes for the feed to and the product produced in the specific test. Therefore, accurate work indices can be obtained from Bond grindability tests when the feed size is finer than standard. To obtain accurate test results, the ratio of reduction should be greater than 6 to 1."

Jan 1, 1972

By Frédéric Flament, Luc Lachance

"The important task of metallurgical accounting that used to be done manually is being more and more automated. Currently available solutions can generate accounting reports without human intervention or almost. However, the practical conditions for fully delivering the promised benefits of those solutions are sometimes overlooked either by end-users or by integrators. The mere fact of filling reports with values satisfying several constraints such as mass conservation constraints does not, by itself, ensure a high quality metallurgical balance. Since managers make decisions based on these reports, it is logical that the metallurgical balance performance be monitored and retrofits be made if necessary. Such retrofits may involve modifications to the measurement strategy or the reduction of the spatial resolution of the material balance model. This requires an estimability and redundancy analysis (ERA) tool to assess the feasibility of the modified model. Since determining by hand the estimability and redundancy of a set of non measured and measured variables is not trivial, even for a medium-sized mineral processing plant, a dedicated ERA tool brings valuable assistance to the metallurgist in charge. The present paper aims first at explaining how a state of the art metallurgical balance is a key enabler for the adoption of a smart process plants paradigm (SPPP) in mineral processing. Then, the fundamentals behind the building of an ERA algorithm are introduced and illustrated by two examples. Finally, the role of an ERA tool in achieving state of the art metallurgical accounting and how it contributes to delivering additional value are discussed.INTRODUCTIONThe Smart Process Plants Paradigm (SPPP)Mineral processors have seen an increase lately in external pressures toward operating performance improvements. Indeed, the globalization of the mining business is a key driver to becoming more productive, more responsive to market changes, to improving corporate governance and reporting standards. Furthermore, mineral processors are also expected by their neighbouring communities to continuously improve their sustainability score. These current pressures differ somewhat from the past ones in leading to a larger than ever number of conflicting constraints that plant managers must concurrently satisfy while maximizing plant profitability. It is in turn not surprising that this context requires actions involving more real-time collaborative actions between more people in more places than ever before."

Jan 1, 2011

By Jean-Pierre Côté, Alexandre Noiseux

"Hatch and ArcelorMittal Mines Canada have combined their expertise to develop a dynamic model of the latter’s Iron Ore Pellet Plant in Quebec, Canada. This model was used to predict the benefits of different expansion projects, as well as asses the best implementation sequence.Two modelling tools were used: a discrete-event simulation software used to develop the dynamic model of the Pellet Plant and an in-house process simulator. The latter was based on operational data and known behavior of the process, and was calibrated with historical data gathered under different operating conditions.The results of the process simulator were used as input values for the dynamic model, which included all of the operation steps of the pelletizing process, from concentrate beneficiation to baking of the pellets. The model has been validated by assessing its capacity to represent adequately the operation of the plant during a 20-month operating period. It was able to reproduce the plant capacity within a precision of 1.2%.Simulated “what-if” scenarios allowed quantifying the production increase associated with the addition of new equipment such as ball mills or balling discs, implemented individually or in combination. Results revealed a potential 10% increase in the annual plant production by adding additional equipment in the two most limiting sectors of the plant, grinding and agglomeration. The simulation results made it possible to quantify the synergistic effect of two projects on plant throughput; the model showed that the combination of two projects resulted in a greater production increase than the sum of the same projects evaluated individually.Now that it has proven reliable, the simulation tool can readily be used by ArcelorMittal Mines Canada process engineers to perform simulations for equipment optimization, and to evaluate the impact of various operating parameters such as equipment availability and tonnage, product mix or concentrate quality. This work showed that a reliable model can be developed based on known parameters in combination with the proper integration of the process knowledge."

Jan 1, 2010

By G. Halverson, K. Blower

"Giant Yellowknife Mines Limited (Giant) has been producing gold since 1948. Located at Yellowknife, Northwest Territories, Canada, see Figure 1, the mine processes approximately 1,200 short tons per day of 0.23 ounce gold per ton of ore.The majority of the gold in the ore is sub-microscopic and associated within the mineral arsenopyrite. This association complicates the extraction method since straight cyanidation will only recover 30% of the gold. The operations at Giant involve four essential steps, with an overall recovery of 87%. The steps shown in Figure 2 are crushing and grinding, flotation, roasting and cyanidation. These processes result in a number of effluent waste streams characterized in Table 1. These streams are combined and sent to a tailings storage pond for subsequent treatment prior to discharge to the environment.Giant is located just north of the 62nd parallel where winter temperatures drop to -40°C for extended periods. This entails storing effluent in dams over the seven month winter period for treatment during the relatively short summer season.The treatment method currently employed at Giant is an ""alkaline chlorination-arsenic precipitation-flocculation"" process. The pond water is subjected to three phases of treatment prior to being discharged to the environment as shown in Figure 3."

Jan 1, 1987

By M. C. Villeneuve

A numerical modelling approach was developed to explicitly simulate geomechanical characteristics of intact rock: mineralogy, grain size and fabric. The approach involved creating a representative constitutive model for each of three common rock-forming minerals: mica, quartz and feldspar. The constitutive models developed are valid within the low confinement realm of excavation boundaries, where tensile fracture processes dominate. The mineral types were assigned to numerical elements, which were associated with each other through an algorithm created in a finite difference model, FLAC 2D (Itasca 2007a), to simulate real crystal geometries and orientations. The numerical models were used in a parametric investigation of the geomechanical characteristics and compared with published observations of the rock yielding process in laboratory testing. This approach has allowed the explicit grain-scale investigation of the impact of geomechanical characteristics on rock yielding at low confinement, leading to an improved mechanistic understanding of excavation-scale rock yielding processes at excavation boundaries.

May 1, 2009

By O. K. Mahabadi

The scope of this study is to simulate the behaviour of a layered rock sample under standard laboratory Brazilian Disc test using an innovative combined finite-discrete element method (FEM/DEM) research code. The influence of layering and the direction of loading on the samples behaviour are studied through various layering and loading configurations. A parametric study including the effect of loading rate on the tensile strength of samples has been carried out. This paper demonstrates the suitability of the numerical approach to explicitly model rock deformation and failure.

May 1, 2009

By Jan W. Merks

"The settlements between mines and smelters are based on the mass of metal contained in concentrate lots. The mass of contained metal is a function of the wet mass of a lot, and its moisture content and metal grade while the variance of the mass of contained metal is a function of the same variables and their variances. Uncertainties in the measurement procedure applied to estimate each variable are quantified in terms of precision and bias. The absence of bias and an acceptable degree of precision are the keys to effective risk analysis and loss control.ISO Technical Committee 183-Copper, Lead and Zinc Sulfide Concentrates, has developed standard methods for sampling and weighing, estimating precision and testing for bias, the determination of mass loss on drying, the calculation of the mass of contained metal and its variance, and various analytical methods for major and minor metals. The variance for wet mass is calculated from the precision of the scale while the total variances of moisture and grade are estimated from pairs of interpenetrating primary samples. The variance of contained metal can be calculated either from mine data or from smelter data but the risk associated with the settlement is largely determined by the absence or presence of bias and the variance of differences between exchange assays.ISO/TC 183 does not deal with precision and bias for exchange assays but is developing guidelines to determine realistic splitting limits. Given that the exchange of assays between mines and smelters accounts for a large part of the risk associated with the settlement, precision and bias for exchange assays should be measured and monitored. The need for umpire services can be reduced by applying statistical quality control (SQC).Calibrating static scales, moisture balances and analytical balances, bias testing of sampling methods and sample preparation procedures, estimating the variance of a single variable or the variance of a function that a set of variables defines, and monitoring precision and bias for exchange assays, are the quintessence of metrology, the science of measurement, when applied in mining and metallurgy."

Jan 1, 1995

By Jan E. Nesset, Phillip J. Mackey

"Studies have shown that there is a correlation between commissioning and start-up time, i.e. time to reach full plant design capacity, and the degree of innovation and technical/engineering uncertainty in the project design. The paper describes the types of start-up curves for plants according to the level of technology or uncertainty in the process design. The need to properly describe and understand the commissioning and start-up time is examined using these curves and the importance on project success is highlighted. A number of case histories of recent mining and metallurgical projects are also reviewed. Finally, the paper reviews how to better assess the level of start-up uncertainty associated with a project and makes suggestions for decreasing the commissioning time due to technological risk.INTRODUCTIOThe start-up of a new mining and metallurgical facility, a much anticipated event, is a critical step in the life of such a project. Achieving a successful start-up and reaching the design capacity on-time requires careful planning, a thorough understanding of the key technological and engineering components involved and a well-trained and adequatelystaffed start-up crew. In this paper, some of the aspects related to start-up of a mining/metallurgical project will be explored, and in particular, the impact of the technology component will be examined. In order to illustrate this, a number of case histories based on recent mining and metallurgical projects will be discussed. Ways of assessing and understanding the uncertainty associated with technological issues related to start-up of a mining and metallurgical plant will also be examined.The history on mining and metallurgy is full of success stories of great mines and mills that are now considered “company makers” (well known examples of such successes include the Horne mine for Noranda which had a 1 year start-up in 1926-27, the Dome mine for Placer Dome, the Grasberg deposit for FMI in Indonesia). Less well documented are specific studies relating to plant start-ups and what constitutes a success or a failure, however there are a few notable references in the literature and these will be reviewed."

Jan 1, 2003

By Michel Ruel

"PID loops – also known as base layers or regulatory controls – form the foundation of any plant’s control layer and it is generally accepted that PID loop health has a direct impact on the plant’s bottom line. Keeping these thousands of control loops in good health has a direct impact on the plant’s bottom line: it reduces variability, improves quality and throughput, and decreases energy costs and raw material usage. This has been well proven and documented in many articles. In the last ten years, technology has evolved: control engineers can now keep track of control loop performance and link it to the root causes of variability. However, they lack industry-wide best practices and criteria for determining what can be considered as excellent, good, fair, or poor performance. To ensure that PID control loops continue to perform well and meet their objectives, control engineers need a combination of (a) an appropriate monitoring technology and (b) best practices for doing the maintenance of base layer controls.Most plants have limited resources and operate with an increasing number of constraints; therefore, using the right tools and including reports in daily workflow is key for success. This paper suggests roles and responsibilities of key players needed to work toward control performance best practices and discuss organisational changes to succeed.INTRODUCTIONAlthough PID controllers are reliable workhorses in the process industry, they are often underrated. Since the 1980’s, computer control has evolved and the control algorithms used by control engineers have become increasingly complex. However, PID controllers still handle over 90% of the control loops in operating plants. Table 1 shows the savings that can be realized by achieving best-practice in several different process control categories (Brisk, 2004). The savings are expressed as a percentage of production costs."

Jan 1, 2011

By R. H. O. Wagner

"This paper describes the bench testing, pilot plant studies of alternative flowsheets, to determine the feasibility of gold recovery, by carbon in pulp extraction, in the Schumacher Mill gold flotation tailings. The economics were calculated based on pilot plant recoveries and actual industry operating costs.The economics of the carbon in pulp treatment of the gold flotation tailings were not justifiable. The economics of desliming the gold flotation tailings followed by carbon in pulp treatment were marginal, but favourable. INTRODUCTIONPamour Porcupine Mines, Limited is a gold, silver and copper producer located in Timmins, Ontario. In 1983 the head grade of Pamour's gold ores averaged nine one hundredths of a troy ounce of gold per short ton.Pamour operates two area mills to concentrate the gold ores. Free milling gold ores are processed at the Pamour One Mine mill, whereas the refractory carbonaceous gold ores, and complex gold, silver, and copper sulphide ores, are processed through the Schumacher Mine mill.The complex Schumacher mill feed ores do not all react favourably to flotation methods, consequently this study was commissioned to investigate the feasibility of the carbon in pulp cyanidation of the Schumacher mill gold flotation tailings. These tailings contain ten to fifteen percent of the gold in the mill feed. During the study period, the average grade of the Schumacher mill flotation tailings was nine one thousandths of a troy ounce of gold per ton, or one quarter of a gram of gold per metric tonne."

Jan 1, 1984

By E. W. Johnson

"The title of my paper is ""I 'Wonder Where The Yellow Went"". As you can readily guess, it is not related to an old-fashioned ""Pepsodent"" sales pitch! Rather, it concerns a situation which arises all too frequently in some gold milling operations, namely when actual ounces of gold production do not come up to estimated or projected ounces. The boss man scratches his head and wonders where the missing gold has gone and how it has disappeared. The problem, if one exists, is amplified if one forgets that an estimate is just that - an estimate - and not a hard fact - even though the numbers are produced by computers. And yet, from the point of view of engineering; financial and production planning and security; it is most desirable to tie together all data relating to production estimates, losses and actual production.In the course of this paper reference will be made to the term ""variance psychology"", to a few of the physical properties of gold and to the nature of its occurrence in Canadian ores, in relation to the topic under consideration. Activities which have a bearing on production estimates will be reviewed. These include: weighing, sampling, assaying, estimating residual losses and, in cyanide plants, gold precipitate production. Mention will be made of apparent, unaccounted for and unknown losses. In conclusion, I shall draw from personal experience and relate three examples of ""Where the Yellow Went''."

Jan 1, 1986