Search Documents

By Donald A. Cranstone

"This paper provides a bird's eye view of the pattern of gold discoveries since 1823, the year of Canada's first gold find, up to 1983. Within this 160-year period, more gold was discovered in Canada during the last ten-year period (1973-1982) than during any previous decade. The record suggests that the potential for further major gold discoveries is high, which bodes welt f or the future of Canadian gold mining.IntroductionCanada has an impressive history as a gold producer. Since 1921 we have occupied second or third place among the world 's gold producing nations, after South Africa and most of the time the U.S.S.R. Canadian gold production over the last hundred years has added up to some 6 per cent of all-time world production.The sharp rise in the price of gold in the 1970s led to a surge of gold exploration, which in turn resulted in the discovery of significant quantities of gold in Canada. How do these recent discoveries rate in comparison with those of the more distant past? For example, are the quantities of gold in notable discoveries of recent year s, such as Hemlo and Detour, in the same league with the amounts of gold discovered around 1910 in the Porcupine and Kirkland Lake district s of Ontario? Can the combined gold content of the recent discoveries be likened to that o f the almost 40 major Canadian gold deposits, along with numerous smaller ones, discovered in Canada during the 1920s and 1930s? What are the implications of the more recent gold discovery record for future rate s of gold discovery in Canada?This paper attempts to glean answers to such questions from an overview of the period that begins in 1823, the year of Canada's first gold discovery. Gold deposits come in two major types: placer deposits and lode deposits in solid rock . Placer gold is distributed in sand and gravel. Gold in solid rock occur s mostly in gold ore filling fissures."

Jan 1, 1985

Background ArcelorMittal?s (formerly QCM?s) Port Cartier iron ore pelletization operations 2 side-by-side induration furnaces each producing 578 t/h of cured pellets Firing No. 6 fuel oil and coke breeze (in bed) Very large heat losses to be addressed primarily by gas recirculation

May 1, 2011

By James H. Savory

"In recent years, there has become a growing interest toward a better understanding of balance rope behaviour. With this in mind, the auth or will present a comprehensive summary of field experience, as well as conclusions. In presenting this overview, it is the intent to stimulate further interest as well as provide a framework from which further in-depth investigation can be initiated.Topics which will be covered include rope cores, lubrication practices, loop dividers, rope failures, etc.IntroductionThe first friction hoist was designed in Belgium in J862 by Lemielle. However, it was not until 1877 when Carl Frederich Koepe installed a friction winder sheave at the Hanover Colliery in Germany, that the design was given serious consideration. The winder consisted of a motor-powered sheave over which a single rope was installed, and the end s attached to two conveyance s. Beneath the conveyances a rope was connected, designed to reduce the chance of slip. This was the first balance rope.This earl y mod el is a far cry from the modern tower or ground mounted Koepe winders employed throughout the mining industry today. There have been vast improvements in all aspects of the Koepe winders. One such improvement is in the balance ropes used. This particular area will be concentrated on, with emphasis on selection, maintenance and field experience.SelectionThe sale purpose o f a balance rope is to counter the head rope weight. Obviously then, in selection o f balance ropes, weight is of primary concern. Usually the head and balance ropes are equal in weight. However, sometimes balance ropes which are 5 to 10070 heavier than the head ropes are employed. These reduce horsepower demands during peak acceleration periods, and assist in overcoming inertia during retardation periods. As well, it is sometimes preferred to reduce the balance rope weight, in order to compensate for the eventual weight buildup of lubricant, dirt and debris."

Jan 1, 1983

By J. E. Lee

"This paper describes the current practices and experiences, and discusses design parameters concerning slurry pumping at Inco Metals Company in the Sudbury District. It deals mainly with the long tailing and concentrate lines. Subject matter is arranged under the following headings: pumps, lines, instrumentation, system design, operation.IntroductionThe Inco Metals Company is the mining and processing section of INCO LIMITED. In the Ontario Division, it operates 13 mines, 5 mills, a smelter, an iron ore plant and 3 refineries, all engaged in the mining and recovery of nickel and copper. This paper describes the slurry pumping practices in the mills of the Ontario Division.The PumpsWe have in our Sudbury area mills most of the available types of slurry pumps, in sizes from one inch up to twelve inches, a total of about 400 pumps. The rubber-lined, horizontal shaft pump is; most numerous. However, for certain applications, the hard-metal pump is essential, and is becoming more popular for general mill service. Two types of vertical shaft pumps are used. There are a few diaphragm pumps on small thickeners, but the larger thickeners have centrifugal pumps either directly connected to the under-flow or with one pump handling several gravity underflows. We have not used piston or plunger pumps in our mills."

Jan 1, 1978

By E. Wright

"INTRODUCTIONThe Ruttan Mine is located twenty-two km east of Leaf Rapids in northwestern Manitoba. The copper-zinc exhalative massive sulphide deposit is comprised of a series of individual and interconnected en echelon lenses which strike east 20 degrees north and dip south 70 degrees. The mine has been owned and operated by the Hudson Bay Mining & Smelting Company since July 1987. Sherritt Gordon Mines Limited began mining the orebody in 1973 from an open pit, but production has been entirely from underground since 1980. HistoryDuring the progression from open pit to underground mining, it was recognized that manual methods of ore reserve calculation were inadequate for responding to changing metal prices and evaluating alternative mine plans. The mine planning system that was developed is described in the paper, ""Computerization of Ore Reserves and Mine Planning Functions at Ruttan Mine"", presented by B.C.W. Haystead at the 1984CIM Annual General Meeting. At that time, programs for core logging, diamond drill hole plotting, grade interpolation, ore outline digitizing, and reporting had been developed by the Systems Department for use on Sherritt's UNIVAC 1100 mainframe computer. Telecommunications linked terminals at the Ruttan coreshack and office to the UNIVAC in Fort Saskatchewan, Alberta. The same mine planning system was also used at a second property.Operation of the mine planning system was expensive, and it was evident that increased use required a costly upgrading of the UNIVAC in order to maintain an acceptable service level to all users. The system was re-evaluated,' and it was concluded that: core logging was not easily adaptable to a new property, nor flexible enough for exploration; plotting, grade interpolation, and digitizing required considerable manual effort; and that the UNIVAC operating system was difficult to learn.It was recommended that a study should determine if: core logging could be redesigned and/or implemented on a microcomputer; interactive graphics be used to reduce the effort for functions requiring visual interpretation; and the creation of input processors would assist users in setting up parameters for guiding the programs."

Jan 1, 1989

By Arnfinn F. Prugger, Fritz F. Prugger

"The Saskatchewan potash beds are overlain by a number of water-bearing formations. There are water zones in the footwall as well. These aquifers created difficulties during the construction of shafts to the potash beds at 1()(}() m depth, and they pose a continuous threat to the safe and efficient operation of potash mines in the province.The first shafts to be sunk in Saskatchewan encountered major problems during shaft sinking. Of the 17 potash shafts started in the province, five had major water inflows during shaft sinking or after production startup, and one was abandoned because of water problems. The biggest obstacle encountered during shaft sinking was a poorly consolidated sand-shale-clay-siltstone unit commonly called the Blairmore Formation, which occurs at a depth of about 500 m. The initial problems with shaft sinking were overcome in the 19605by using European methods of freezing and the installation of cast-iron tubbings or a double-steel lining through water bearing horizons.From the start of production in the early 19605, operators have been aware of the ever-present hazard of potential water inflows into mine openings from permeable units in both the hanging wall (carbonates of the Dawson Bay Formation which occur 15 m to 50 m above the mining level) and the footwall (carbonates of the Winnipegosis Formation which occur 50 m to 150m below). This awareness has been reflected in conservative over-all extraction rates and mining layouts specific to local conditions. In spite of this consciousness, a number of water problems have occurred in operating mines. These have resulted in everything from simple increases in expenditure to the shut-down of one mine in 1987.In this paper a history of water problems in the Saskatchewan potash industry is given, followed by a summary of measures that are now commonly used to avoid these problems. A description of each inflow event is given, and the attempt is made to analyze the causes of each occurrence. Based on what was learned f rom each event, suggestions are made to prevent future water inflows into potash mines."

Jan 1, 1991

By L. Sun

In this work, strontium doped lanthanum ferrite (LSF) and their composite cathodes were fabricated by screen printing on yttria-stabilized zirconia (YSZ) electrolytes. Their electrochemical properties were evaluated by using electrochemical impedance spectroscopy and steady-state polarization methods. All the cathodes were screened by comparing their total polarization resistances before and after activation processes at different temperatures. The improved LSF composite cathodes were also screen-printed on YSZ/NiO half cells to fabricate solid oxide fuel cells (SOFCs), which have reached the output power densities of 170 mW/cm2 at 650 °C, 230 mW/cm2 at 700 °C, 304 mW/cm2 at 750 °C and 388 mW/cm2 at 800 °C.

Jan 1, 2005

By D. S. Evans

Major Project Blues Over the last two decades, almost every major project has suffered significant cost overruns, significant schedule delays, and/or poor quality work. Exceptions are rare and most projects that tout being ?On Time and Under Budget? have re-cast schedules, significantly modified scope or revised base estimates. The problem isn?t new. In the late 1980?s the Rand Corporation published a study showing that 80% of all major projects had significant cost and schedule overruns. The problem is not restricted to the Canadian environment, or to a specific industry. It is a global problem.

May 1, 2013

By T. Negeri, M. Boisclair

"A hybrid process consisting of flotation and magnetic separation has been developed to concentrate multi-phase rare earth minerals associated with a carbonatite ore that contains a significant amount of niobium. Fine grinding to a d80 of <65 microns was required to obtain acceptable separation of 92% of the contained total rare earth oxide (TREO) and niobium in half the feed mass. Whereas it was possible to obtain 85% of TREO in 28% of the feed mass through staged cleaning, it was demonstrated that at least 87% and 85% recoveries of TREO and Nb respectively could be achieved in 16% of the feed if the rougher-scavenger concentrates were calcined, quenched, reground to <25 µm and then magnetically upgraded. The paper discusses the overall concept of the flowsheet and the experimental strategies that led to this progress. INTRODUCTIONOut of the approximately 250 minerals known to contain rare earth elements (REEs), only about 60 of them contain REEs in significant amounts. Over 95% of the world’s rare earth oxides (REOs) are believed to occur in the form of three minerals (Pradip & Fuerstenau, 1991) namely bastnaesite (Ce,La,Y)CO3F, monazite (Ce,La,Nd,Th)PO4 and xenotime Y(PO)4. An estimated 51.4% of global REO resources are hosted by carbonatite deposits with the same three minerals as major constituents (Weng et. al., 2015). Any one of the REEs or commonly all of them can be found in these minerals (Ravnaas, 2010). As the demand for critical REEs increases due to the ever increasing applications in advanced technologies, many deposits containing other forms of REE minerals will have to be developed. REE minerals are extensively dispersed and intergrown with several oxides, silicates, carbonates, and phosphate minerals. Consequently it is possible to liberate them only at fine grinding and is thus difficult to separate and produce high grade rare earth mineral concentrates. Most of these minerals contain other important elements such as tantalum (Ta), niobium (Nb), uranium (U), and thorium (Th) in varied quantities. REE minerals in the same ore may exist as carbonates, fluorocarbonates, niobates, silicates, phosphates etc. in association with gangue minerals belonging to different mineralogical groups."

Jan 1, 2016

By S. Banisi, J. Marois, A. R. Laplante

"This paper compares the behaviour of gold and ore in the grinding circuit of Hemlo Mines Ltd. Particular attention is given to grinding kinetics, classification behaviour, and liberation. For grinding kinetics, laboratory grinding of gold flakes and silica, 840 µm to 1200 µm in size (14/20 mesh), was also performed to provide a standard for comparison and to study in a controlled environment the mechanisms of breakage, folding , smearing and impacting.Grinding kinetics were characterized at laboratory scale with breakage and selection functions. The breakage function showed that 75% of gold disappearing from the 840 µm to 1200 µm class reported to the 600 µm to 840 µm; this figure is much higher than for silica (46%) or most other minerals, and implies that comparatively fewer fine fragments are produced from the breakage of gold flakes, on account of gold&apos;s malleability. The selection function of silica was four times that of gold. At plant scale, the ratio of the selection function of ore and gold increased from 6 for at 50 urn to JOO pm to 20 at 500 um to 1000µm.Observation of original flakes and their fragments with the scanning electron microscope showed that the main mechanism of ""grinding"" from the 840 µm to 1200 µm to the 600 µm to 840 µm class was folding rather than actual breakage. This was confirmed by a measure of individual flake weight. The postulated breakage cycle is that relatively globular flakes become increasingly serrated through grinding; they can then regain their original shape through folding and reflattening into cylindrical and spherical shapes, as the majority does; however, some fragments will at this point undergo actual breakage. Photograhic evidence is presented to support the proposed mechanisms.Smearing of gold unto silica was observed when silica was ground after a gold, and embedding of silica into gold flakes when gold was ground after a silica. Both mechanisms occur whenever gold and harder minerals are jointly ground, and result in ""unliberation"" of gold. It is postulated that smearing of gold onto other minerals plays an important role in explaining how gravity recovery can increase over-all gold metallurgy, especially when used ahead of flotation.Classification of gold in the primary hydro cyclones takes place at a much finer cut-size than ore, 20 µm vs 57 µm. This and the slower grinding kinetics result in high gold circulating loads in the primary loop, 6700%, and a gold content in the primary cyclone underflow (PCUF) is twenty-four times that of the ore. The gold size distribution for primary cyclone overflow is much finer than that of the ore, although they are essentially similar in the PCUF.Liberation, or more accurately ""gravity recoverability"", was assessed with a 7.5 cm (3 in.) Knelson concentrator. It varied between the thirties for the primary mill discharge and two cyclone overflows, to highs of 89% and 92% for the PCUF and secondary mill discharge."

Jan 1, 1991

By Orlando Rojas, Javier Pizarro, Romke Kuyvenhoven, Francisco Soto

"The concept of geometallurgy has been extensively mentioned and used in the mining industry in the past years. However, each mining firm has a different understanding of what geometallurgy means, and its actual value varies significantly from site to site. Even in the case of properly constructed geometallurgical models, it is fairly common to observe a sub-utilization of the model in the actual operation, because of its low production forecasting capability on the short term. This disconnect between the geometallurgical model and operations is typically caused by a lack of tuning between the original model and the actual plant operation. Also, the absence of procedures and other standards for reconciliation, affects the confidence and the predictive quality of the model. The geometallurgical model should consist of three different levels: (I) adequate capture of information, (II) data analysis and modelling, and (III) development of the block model. Once the geometallurgical in-situ model is finalized, a process model should be developed of the plant at industrial scale, in order to scale up the in-situ geometallurgical parameters from their base or standard condition, to the actual operating reality of the plant. Finally, the stage of reconciliation will close this process, assuring that any deviation between the geometallurgical model and the actual plant operation is not due to lack of tuning, but due to operational issues such as unattended maintenance requirements or poor process control in the operation. The methodology as described in this paper has been applied at several large mining sites in Chile. The paper presents the characteristics of the proposed methodology, and quantitatively describes its benefits with examples from actual operations.INTRODUCTIONIt probably is fair to say that geometallurgy became widely used in greenfield projects in the Andes region during the mining boom of the 90’s, when massive orebodies were studied and developed, and upfront metallurgical characterisation proved its use for accurate economic evaluation including estimates of plant throughput, energy consumption, final product quality and overall recovery. A few examples of such greenfield studies for concentrator plants in the 90’s are Minera Los Pelambres (Antofagasta Minerals, Chile), Minera Escondida, Laguna Seca (BHP Billiton, Chile), and Rio Blanco (then Montterrico Metals, Peru). Large size hydrometallurgical operations such as Cerro Colorado, Quebrada Blanca and Lomas Bayas (all in Chile), also recognised early on the importance and value of relating ore mineralogy, geological characterisation and metallurgical behaviour."

Jan 1, 2016

By H. J. Oterdoom

SMS Group (SMSG) has always been concerned about safety aspects of electric smelter plants for the ferro-alloy and non-ferrous industry. The awareness for safety increased significantly over the past decades and also SMSG contributed with its design features to a safer operation. Today, safety aspects are not only part of the smelter design but also of the overall plant layout. Explosions in electric smelters are rare, but have a huge impact on peoples’ lives, on equipment and on production. For this reason, SMSG wanted to have a better overview of explosions within the smelting industry in order to see if useful statistics could be extracted. Generally, there is a lack of statistics regarding explosions in the electric smelting industry, and not enough information is collected or published. A significant portion of the explosions seems to be water related. Many companies, both operating and engineering, are trying to handle the risks of water, and apply methods to detect water leakages. Installing such a system is a management decision and management has then to make sure the system is properly implemented. The disadvantages and limitations of these systems should be carefully considered and communicated to operators and authorities.

Jan 1, 2015

By V. Prasad, A. Afacan, K. Popli, S. Dubljevic

Froth flotation is one of the most common and widely used separation processes in the mineral processing industry; however, its complexity means that complete understanding of the process has not yet been achieved. A dynamic fundamental model is needed to enhance the understanding of the process and to improve process control. It is challenging to capture all the phenomena involved; hence, we propose a multiscale modeling approach with a framework ranging from macroscopic (equipment scale) to microscopic (attachment, detachment) models along with distribution models for ore particle sizes. Real-time froth properties are obtained using a vision-based sensing system. These measurements can be used for developing a soft sensor for grade using statistical methods such as PCR (principal component regression) and PLS (partial least squares) regression. The fundamental model is updated in real-time using extended Kalman filtering. Attachment dynamics are studied for different particle size distributions. The modeling and soft sensing is tested on batch flotation cell for the pure mineral galena with different particle size distributions using Xanthate and MIBC as collector and frother, respectively. The model, online measurements and soft sensor are useful in developing monitoring methods for froth flotation.

Jan 1, 2015

Once applied, EarthZyme reduces the tendency for a road to exhibit common maintenance issues: wash-boarding, rutting, heaving, potholing and dust.

Aug 1, 2013

By J. Kalejta

Ever increasing depth of mineral extraction presents a challenging environment for hard rock underground mines. High in-situ stresses and associated seismicity with potential rockburst hazards are the major decisive factors contributing to the choice of a ground support regime. Conventional ground support systems, designed primarily for static loads, are not always capable of providing safe working conditions for underground personnel in seismically active mines. Systems specifically developed to resist dynamic loading and allowing for larger deformations are therefore preferred alternatives. High-tensile chain-link mesh has a proven record of successful use in open cut operations in various rockfall barrier installations due to its high energy absorption capacity. It has also been used in underground operations in various parts of the world. This paper describes a method of mechanized installation of a chain-link mesh evaluated at the Kiruna mine in Sweden and South Deep Mine in South Africa. A purpose-built mechanized roll mesh handler was used in this evaluation. The handler is compatible with all standard multi-boom underground drill rigs and is operated utilizing the hydraulic circuit normally used for the feed arrangement. A number of key performance indicators have been identified as success criteria: ??A 30-50% faster installation compared to the conventional methodology ??The mesh should sustain any blast damage when applied to the blasting face ??No unraveling of the mesh when cut ??No failure of mesh under normal conditions Although the energy-absorbing mesh combined with yielding rock bolts is a ground support system of choice for rockburst prone conditions, the operational upsides recognized during the evaluation imply that this product can also be competitively used in less demanding ground conditions where customarily weld mesh or shotcrete are used as a primary surface support. The main conclusion: a fast, safe and fully mechanized installation.

Aug 1, 2013

By Ed McConaghy, P. LaRochelle

The Ambatovy Nickel Project comprises the development of mine, hydrometallurgical processing facilities and necessary infrastructure in Madagascar, positioning Ambatovy as one of the world’s largest producers of nickel and cobalt from lateritic ore. SNC-Lavalin successfully provided comprehensive Engineering, Procurement, Construction Management (EPCM) services for this major capital project which, apart from its sheer size, scope, and engineering design, represented significant location, infrastructure, labour, environmental and climatic challenges. Today, the project has reached 90% of its nameplate capacity and is in its final performance optimization stage. This paper will discuss the significant project management effort that was required to successfully address the complexity of this major project. It will cover the main project management control systems and the engineering tools and processes used during the detail engineering, procurement, and construction phases to achieve optimum project execution.

Jan 1, 2015

By Yanping Xiao, Ton van Sandwijk, Gus Van Weert, Larry McHugh, Thomas Zabev

"All attempts to remove significant amounts of arsenic from enargite via neutral or reductive pyro routes have required operating temperatures between 600 and 700 º C, too high to propose an equipment development program. Operating at reduced pressures, or using hydrogen or elemental sulphur as a sweeping gas did not help to lower these temperatures. Admixing iron powder gave excellent results, but still required 700 º C. By fortuity, in recent experiments to produce tennantite from an enargite mineral sample, it was observed that enargite was completely converted to bornite, chalcopyrite with minor residual tennantite within an hour at 400-450 ºC. This is believed to be due to the presence of pyrite in the mineral sample treated. All earlier decomposition work appears to have been done on predominantly enargite (i.e. low pyrite) samples, or at high operating temperatures, ignoring possible reactions at lower temperature. The decomposition of enargite into iron substituted tennantite plus some bornite under evolution of liquid elemental sulphur, followed by further decomposition of this tennantite into bornite and volatile arsenic sulphides, appears experimentally feasible at these low temperatures and mimics described ore genesis pathways at > 250 ºC in saline solutions. We suggest two pathways to describe the observed experimental results:20 Cu3AsS4 + 12 FeS2 ???5 Cu10Fe2As4S13 + 2 Cu5FeS4 + 31 S(l)31 S(l) ???31 S(g) (evaporation)5 Cu10Fe2As4S13 ???10 Cu5FeS4 + 10 AsS(g) + 5 As2S3(g)"

Jan 1, 2014

By Eric Best

"In the past two to three years, all continuous mining machines at the Potash Corporation of Saskatchewan Mining Limited, Lanigan Division, have been changed over to a new cutting pattern using 1 -inch- and 1-3/l6-inch-diameter self-sharpening pencil bits. The bits are arranged in a cone-type pattern across the face on 2 1/2 -inch and 3-inch spacing. The result is that the ore is removed by chipping and flaking. Previously, rap-lock bits were used, with the pattern being several bits in single rows at wide spacings. The cutting action was grinding and breaking.Besides improved penetration rates and lower bit costs, there are many other advantages. The ore size is much smaller, resulting in less wear and tear on conveyors and other associated equipment. Feeder-breakers are no longer routinely set up at the head end of every extensible belt, as the banana-shaped chunks of ore that used to require breaking are no longer produced. Also, less dust is generated by the cutting heads and fewer stops are required for bit changes. Even the mill operation has been improved, as fewer fines are produced.Further improvements are still sought after, but, all in all, the new cutting pattern is considered to be a giant step forward from the old method."

Jan 1, 1981

By G. R. Waters

"The technology behind industrial applications such as Hydrometallurgy continues to evolve, as do the corresponding regulations and specifications that ensure the safety of workers and the environment. Valves make up key aspects of this technology, and often represent the single point of failure for entire operations. More often than not, the valves being used in Hydrometallurgy applications are known as Severe Service Valves (SSVs) due to their ability to withstand the extreme conditions. Most experts agree that SSVs are identified by applications, and that these applications are challenging to the valve’s ability to provide a minimum acceptable level of performance over a minimum acceptable duration. The Manufacturers Standardization Society (MSS) of the Valve and Fittings Industry has designated Ross Waters, President of CGIS, as Lead to the Task Force put in place to determine the minimum requirements a valve needs to be able to perform to when faced with extreme conditions, whether from pressure, temperature, toxicity, solids, or usage. This paper serves to determine what the minimum requirements are for SSVs in Hydrometallurgy applications. It will draw on five decades of industry experience from one of Canada’s leading valve experts.INTRODUCTION Severe Service Valves Severe Service Valves (SSVs) mean different things to different people. Until today, defining SSVs had little if any global agreement or recognition. That is about to change as the Manufacturers Standardization Society (MSS) has accepted an application to produce a Standard Practice so defining them. This paper provides information about the selection of SSVs in all industries but focusses on metallurgical processes and applications and offers examples to illustrate the successful and unsuccessful use of these valves. All of this with the purpose of raising the awareness of the industry on all sides, from the suppliers and manufacturers, specifiers and to the users and owners of them. It also supplies tools to understand where and why to separate SSVs from general purpose or commodity valves (GPVs). SSVs are often identified by applications, and these applications are challenging to the valve’s ability to survive. Within these applications the elements that make the service severe are being analyzed, quantified and qualified. From this, we expect to offer objective and repeatable definitions and guidance to improve the experience of SSVs, reduce unnecessary costs, provide longer service life and process runs, improve safety and reduce environmental issues."

Jan 1, 2017

By M. Bourassa, P. Roy

"This paper describes the automation of Kiena's grinding circuit which comprises an open SAG mill circuit and a closed ball mill circuit with two stages of hydrocyclones. The circuit has been instrumented according to a two level control strategy.The regulatory control strategy is implemented in an industrial controller. This controller does data filtering and calculation (flowrate and solid content) and regulates the pump box levels (2), the water additions: or the pulp percent solids (4) and the fresh ore feed rate.The supervisory control calculates for the industrial controller, the fresh ore feed rate setpoint and the pulp percent solids setpoints in order to optimize the grinding circuit operation. It is made of two modules, one for data filtering and one for control optimization. The filtering program uses modern tools to adjust the measured value and estimate missing data (dynamic material balance, on line estimation of model parameter, etc.). The control program acts as an expert system which combines the benefits of model-based control and rule-based control. A multivariable predictive controller (DMC) is used to maintain the circulating load and the hydrocyclones feed percent solids (which is correlated to particle size) while the rule-based control validates the previous output, detects and avoids the SAG overloads and sump pumps overflows.The system has been operating since MAY 1992 and the benefits obtained are discussed.INTRODUCTIONIn 1990, Les Mines d'Or Kiena decided to go ahead with a project for automating its grinding circuit as proposed by the Centre de recherché minerals (CRM). Its objectives in doing so were to marginally increase production capacity (1 % ) and to create a new challenge for plant staff. The main stages of the project, which involved CRM, Kiena Mines and a consultant, are shown in Table 1. The project was completed in May 1992 with the implementation of a supervisory control system. This text describes the main features of the system with a special focus on its most innovative aspects."

Jan 1, 1993