Search Documents

By Peter Mitschke, Friedrich Rosenstock, Heinrich Meiler

"The development of the Centrifugal Mill jointly by Chamber of Mines (Johannesburg) and LURGI GMBH (Frankfurt) had been described by L. Kitschen e.al. during the 14th IMPC at Toronto in October 1982. Further investigations, using a 55 kW semi-industrial mill, installed at LURGI's Research Laboratories, were concentrated on efficiency and wear in grinding.The test results showed, that efficiency increases when increasing the ball load between 60 and 80 % by volume. Furthermore, the efficiency is increased by speeding up the mill, thus increasing the centrifugal force to up to 14 times the force of gravity. Applying these findings to grinding of a Canadian iron ore spiral concentrate, a saving in grinding power of 30 % was experienced, when comparing with ordinary wet ball milling.Investigations of several shapes of the grinding chamber cross section and linings and several types of materials showed, that no increase of liner or ball wear will occur compared to conventional ball mills when choosing suitable lining patterns. 1. DEVELOPMENT OF THE CENTRIPUGAL MILLThe centrifugal mill was developed with the aim of finding a grinding unit that can operate at a higher power concentration than the conventional tumbling mills.LURGI felt the necessity for such a mill in connection with a hydrometallurgical process developed at the beginning of the seventies, which incorporated an activating grinding step as one of the essential process features. At that time, the Chamber of Mines in Johannesburg was investigating possibilities of underground wet grinding and preconcentration of gold ores aiming at a decrease of mining costs by reducing the quantities, especially for mines operating at depths up to 3600 m, by changing over from rope hoisting to hydraulic transportation systems. After the Chamber of Mines had initiated the development work on its own, in 1974 a cooperation between the two organisations was started."

Jan 1, 1985

By A. D. Misener, L. G. D. Thompson

"IntroductionIt was shown by work previously reported that reliable geothermal gradients representative of various regions of the earth's, surface could he obtained from extended surveys of the rock temperatures in operating mines. These surveys showed the presence of anomalies in the isothermal levels and the existence of regions of regular behaviour. Below about 3,000 feet the temperature gradients became quite regular over the areas stud1ed. It is these regular gradients which are taken as representative of the region, wherever possible.The present paper describes the extension of this work which was undertaken to obtain further temperature gradients with a view to correlating them to other geophysical properties determined for the same regions. The additional data obtained have indicated possible reasons for the temperature anomalies. The experimental procedure was similar to that already described and was again made possible through the wholehearted co-operation of the management and the assistance of the personnel of the mines concerned. The research was entirely supported by a grant from the Research Council of Ontario.The equipment was a distinct improvement over that previously used and made it possible to take many more accurate temperature readings in a given time. Thermistor resistance thermometers were used, encased in silver sheaths and protected against damage by heavy copper casings. These were calibrated with a standard platinum resistance thermometer.The bridge used with these thermometers was one of a number designed and built by the Meteorological Service and designated M.S.C. Thermometer Indicator Type C-3. This very convenient and portable instrument operated from batteries in the instrument case. Multiple thermometer connections were provided so that rapid readings• of a number of thermometers was possible to accuracy of the order of 0.1°F. All electrical connections on the panel are covered with Dow-Corning ignition compound to prevent; electrical leakage due to moisture condensation. This, together with the high resistance of the thermometers and components of the bridge, gave much more reliable results than previously obtainable."

Jan 1, 1950

By X. Grant Chen, A. R. Farkoosh, M. Pekguleryuz

Nano-scale Al-(Fe,Mo)-Si dispersoids form during solution treatment in grain interiors (intradendritic regions) of cast Al7Si0.5Cu0.3Mg alloyed with 0.3Mo (wt.%). Unlike the Q-Al5Cu2Mg8Si6 and ?-Al2Cu age-hardening precipitates which coarsen resulting in loss of strength in Al-Si-Cu-Mg alloys, these Mo-containing dispersoids are thermally stable and retain their strengthening effect at elevated temperatures. Consequently, the age-hardened Al7Si0.5Cu0.3Mg0.3Mo (wt.%) exhibits high strength and creep resistance at 300 °C.

Jan 1, 2014

By James A. Finch, Stéphanie Gélinas

"A colorimetric technique, originally developed for methyl iso butyl carbinol (MIBC) based on the Komarowsky reaction, has been extended to a wide range of common mineral flotation frothers. The procedure, which is briefly described, is shown suited to on-site determination in a series of industrial campaigns. Some frother observations are described, including: concentration evolution down a bank, distribution between pulp and froth phases, and a mass balance around a circuit. Minute amounts of frother in recycle water were also measured.The contribution of frother and alcohol by-products in the analysis was generally additive for plants using aliphatic alcohol type frothers and decoupled for systems using glycols. Typically, U250C (polyglycol) was used at higher concentration (>10 ppm) than MIBC (aliphatic alcohol) (<10 ppm) or F-150 (polyglycol) (<3ppm).The ability to monitor frother easily will help interpret gas dispersion behaviour and should permit further optimization of flotation cell hydrodynamics and chemistry. The capability to directly measure should help optimize the type, amount and location of frother addition, and to assess frother recycled with reclaim process water.INTRODUCTIONOwing to the lack of analytical methods, the task of determining frother concentration in plant flotation circuits is not routinely performed. Industrial sites seldom outsource analyses that are both costly and inconvenient, due to the necessity of shipping freeze-dried samples to remote testing facilities. A more direct colorimetric method has recently been demonstrated for the laboratory characterization of common industrial frothers (Gélinas and Finch, 2004). The approach, based on the Komarowsky reaction (Penniman et al., 1936; Coles et al., 1942; Duke, 1947), was previously described by Parkhomovski et al. (1976) for methyl iso butyl carbinol (MIBC). The reaction involves the interaction of the OH group of a higher alcohol, concentrated sulphuric acid, and salicylaldehyde, to yield coloured solutions with intensities that can be analyzed by UV-visible spectrophotometry."

Jan 1, 2007

By Yan Yang

The Jinfeng gold mine is one of the three gold mines that Eldorado Gold has in China and it is the second largest gold mine in China. The mine is located in a relatively remote and mountainous area in the southwestern Guizhou province. The orebody is similar to the Carlin type deposit in Nevada. The mill throughput was designed to treat 3,600 tones per day. The ore comes from open pit mining and underground mining, and is fed to one-stage jaw crushing. The crushed ore is stockpiled, reclaimed and then fed to the SABC grinding circuit. After classification with hydrocylones, the ore is floated, and the resultant concentrate is oxidized biologically in a process known as BIOX®. The oxidized concentrate is then washed in a CCD circuit and finally leached with cyanide in a CIL circuit for gold extraction. Jinfeng produces about 180,000 oz of gold per year. This paper introduces operations and modifications of SABC circuit over the past three years, and then describes some technical characteristic of the SAG mill and other circuits in the processing plant. Major modifications and improvements on the SABC circuit are described in the paper, including: 1. Impacts of grinding ball and mill liner with respect to material selection, quality, hardness, manufacturing methods, and interactions between mill liner and grinding ball. 2. Effect of milling operation parameters on milling efficiency. 3. Impacts of SAG mill scats and modifications of grate size to improve mill throughput. 4. Design and construction of a pebble crushing system, and recycle the crushed scats to the SAG mill to improve mill throughput. With the above-mentioned modifications, 20% increase over the design in mill throughput has been realized over the past three years, and operating cost has been significantly reduced. However, some issues still remain, and thus process modifications and technical innovations are still needed.

Jan 1, 2012

By Hua-Zhong Yu

Gold is one of the most widely used substrate materials for self-assembly. Self-assembled monolayers (SAMs) prepared from alkanethiols and their derivatives on gold are the most reliable and well-studied systems to date, which have attracted much attention because of interest in two-dimensional molecular assemblies, and of their potential applications in molecular devices, sensors, and functionalized interfaces. Gold electrodes are also very important in microelectronics and electroanalytical chemistry. This work reports the characterization of a new type of gold substrates for self-assembly. The gold substrates are constructed from recordable compact disks (R-CDs) with simple and straightforward wet-chemical treatment, and can be made at desired sizes and shapes that satisfy different experimental needs. In particular, an easy, inexpensive, and reproducible method to prepare self-assembled monolayers (SAMs) on CD-R gold substrates after removing the protective polymer films with concentrated nitric acid was described.

Oct 1, 2003

By G. A. Young

THE Geological Survey was created in 1842 to aid the development of Canada&apos;s mineral resources. Although the institution has been in existence for nearly 100 years and has accomplished an immense amount of work, the need for its services remains as great as ever and, indeed, it is difficult to imagine a time when the country will no longer require a Geological Survey. This is so because of the tremendous effort required to make an adequate geological examination of so large an area as is embraced by Canada and partly because what today may seem an adequate geological examination proves, later on, to be inadequate. The ultimate objective of the Geological Survey is to complete the study of all phases of the geology of Canada, that is, to determine how and when the physical features originated, how and when the soils and other unconsolidated deposits were formed, and how and when the solid rocks and associated deposits were produced and assumed their present conditions. When the study has been completed, any only then, will it be possible to make a reliable estimate of the country&apos;s mineral resources. This appraisal embraces more than determining the value of already discovered deposits: it also involves attempting to establish what kinds of undiscovered deposits may be present and where they may occur. The estimating of the potential mineral wealth can be clone only after the geology of the country is known. Not until then will it be possible to forecast what types of valued deposits may occur in a district and what types will almost certainly be lacking. A thorough knowledge of the geology of the country is desirable not only to permit estimating the potential mineral resources but also as an aid in locating and developing mineral deposits whether these are of the nature of a bed of limestone, a seam of coal, a reservoir of petroleum, a vein, an irregular body of sulphides, or any other type of deposit.

Jan 1, 1940

By L. Parthasarathy, S. K. Pan, M. P. Srivastava

"Total iron ore reserves in India, as reported by the Indian Bureau of mines, are 23.6 billion tons. Hematite iron ore reserves constitute about 45 % of the total reserve. If we go by the National Steel Policy-2005 projections, India’s supply of hematite iron ore would not last more than 40 years. Laboratory investigations were carried out for the development of a process flow sheet for treating low grade as well as Banded Hematite Quartzite (BHQ) iron ore. The purpose of these investigations was to aim for better resource utilization.Beneficiation studies were conducted on samples of low grade and siliceous iron ore from the Dalli mines, a captive mine of the Bhilai Steel Plant and BHQ ore from Southern states of India. The process flowsheet, developed for upgrading of low grade iron ore, incorporates wet screening of the ore at 0.5 mm, followed by grinding of the +0.5 mm fraction to below 0.5 mm. The whole –0.5 mm fraction, after grinding, was treated in a conventional hydro-cyclone and stub cyclone followed by treatment in Wet High Intensity Magnetic Separator (WHIMS). This resulted in a concentrate product having >66% Fe content from a feed containing 55.5% Fe with an iron recovery of 48 %. The process flowsheet developed for treating BHQ ore in the laboratory resulted in an increase in iron grade in the concentrate from 43.4 to 61.3 Fe % with a yield of 51.7 % .INTRODUCTIONThe classification of Indian iron ore resources have been made compatible with United Nations Framework Classification (UNFC), which is more scientific and has been adopted in most countries in the world. Iron ore reserves and resources estimated using the UNFC classification system, as of April 01,2000 are about 12,906 million tons of hematite ore and 10,628 million tons of magnetite ore. The Indian Bureau of Mines (IBM) has given a breakdown of these reserves by designation as either lumps or fines of hematite iron ore (Table 1). The percentage of the reserves reported lumps and fines are 42 and 34% of the total reserves respectively. High grade iron ore is less than 10% of the total reserve."

Jan 1, 2008

By Ernest Baafi

A simulation model RoadSIM using the ARENA modelling system has been developed to evaluate roadway development options for a given development panel configuration. The RoadSIM simulation system has provisions for variability/randomness in operations including cutting and loading at the face, tramming and discharge of coal to the boot end and availability of equipment. The modelling system provides a means for assessing the operational limitations of roadway development practices at a particular coalmine. It is a ?what if? tool that allows a range of equipment, configurations and operating practices to be assessed in terms of achievable advance rates and equipment utilisation. Output from the system is in the form of a dynamic visualisation as well as summary reports and detailed logs of operations over time.

Oct 1, 2009

By Graham A. Karklin

"The Echo Bay Mines silver property is located 875 air miles north of Edmonton, Alberta. The mine is on the east shore of Great Bear Lake at Port Radium which is twenty six miles south of the artic circle, in the North West Territories.HISTORYThe area was first surveyed by J. Mackintosh Bell of the Geological Survey of Canada in the early nineteen hundreds. In May 1930, Gilbert LaBine staked claims at Port Radium called ""Cobalt"" and ""Cobalt I"". What he had discovered was an extremely rich vein of pitch-blend. These claims were the birth of mining in this area. The first ore shipment of pitch-blend was hand-picked and flown out of Port Radium in 1931. LaBine's discovery of pitch-blend became of world-wide importance with the outbreak of World War II. In 1939 the splitting of the uranium atom brought this area into world focus. It is generally believed that it was ore from this area that was used in the production of the devastating bombs which were dropped on Japan.It was at Port Radium that the leaching process for uranium extraction was pioneered. The resultant methods have been modified and adopted for use around the world. The mine was originally owned by Eldorado Gold Mines Limited. Echo Bay Mines leased the property in 1964, subsequently purchased it, and have been producing silver and copper every since."

Jan 1, 1981

By Tassos Grammatikopoulos

The St. Lawrence deposit is a large wollastonite skarn (ca. 9 Mt @ 41.3% wollastonite) adjoining the gabbroic-to-syenitic Leo Lake pluton in the granulite facies of the Frontenac Terrane of the Grenvillian Central Metasedimentary Belt. It occurs within a horseshoe-shaped horizon of quartzite, open to the east. The skarn extends over 1.2 km2, and comprises wollastonite-dominant, wollastonite-clinopyroxene, and diverse quartz-feldspar-sulphide (-titanite) layers, interbanded with, and strongly folded within, the quartzite. As in the world-class wollastonite deposits of the Willsboro district, New York State, no carbonate protolith is preserved and calcite is a negligible constituent of the mineralized zones. Wollastonite development is nonetheless ascribed to metasomatism of a metacarbonate unit by magmatogene, silica-rich, CO2-poor (XCO2 < 0.05) fluids, at P = 350 MPa and ca. 560 oC, or, alternatively, 550 MPa and ca. 630 oC. Several high grade zones, averaging >40%, but locally exceeding 80%, of translucent, white, tabular wollastonite grains, with lengths of 0.2 - 2 cm, have been delimited by drilling. Preliminary beneficiation studies yield high aspect-ratio and high-purity (low Fe and Mn) concentrates. St. Lawrence compares favourably to other known prospects and operating wollastonite deposits; having good chemical and physical properties, low iron content, and high aspect ratios (>20:1). In addition, the deposit has a favourable size and location, and will benefit from a low Canadian dollar. Canada currently has no wollastonite production.

May 1, 2002

By K. Norris

"The Campbell mill has undergone numerous changes over the past 43 years. However the rate of change has increased significantly in the last 4 years. A new grinding circuit, carbon-in- pulp circuit, pressure oxidation plant, and column flotation circuit are all new additions. Other changes have occurred in the last 2 years including a large reduction in sodium cyanide consumption, and the introduction of a computer controlled grinding circuit. More improvements are planned in the next 2 years. Some of these improvements include the construction of an effluent treatment plant, and the replacement of filtering and Merrill-Crowe precipitation with carbon-in-leach and electrowinning.Recent changes and improvements to the Campbell mill are a commitment by Campbell mill personnel and Placer Dome Inc. to decrease operating costs, increase gold recovery and gold production, improve industrial hygiene and safety, and meet environmental compliance with current government legislation.INTRODUCTIONThe Campbell Mine has been and continues to be one of Canada's premier gold mines. Campbell Mine is 100% owned and operated by Placer Dome Incorporated. Milling of ore commenced in 1949 at a rate of 300 tons/day producing 58,000 ounces per year of gold. In 1992 Campbell Mine will mill 1340 tons/day of ore at a grade of 0.6 ounces/ton, producing 275,000 ounces of gold at a cash cost of $205.00 Cdn./ounce.This paper discusses some of the changes that have been made to the Campbell mill in the past 4 years. The flowsheet has undergone extensive changes and will continue to change rapidly in the next 2 to 3 years. A simplified flowsheet of the Campbell mill is shown in figure 1."

Jan 1, 1992

By Buse Florian

This paper presents the specific features of the large-scale production and transportation systems (LSPTS), describing the main notions and characteristics of the LSPTS elements. A methodologically unitary approach of the formulation, formalization and solution of the problems regarding the planning of the LSPTS structure development is offered, based on the system description by the aggregation-decomposition method, as weil as on the complex of interreloted models for the planning optimization. The paper describes the characteristics of the complex of dynamic models for investments, utilized in solving the planning problems of the LSPTS structure development.

Oct 1, 2009

By Normand Champigny

The Forces of Change ? The current business environment ? What can we expect for the future on management of physical assets ? The impact of maintenance on production costs ? What is RCM is doing is improving business performance ? The beginnings of RCM The RCM II Process ? RCM II is a structured process that involves employees in determining the best failure management tactics for systems and equipment in their operating context ? The method is an effective approach to risk and cost reduction as well as failure management

May 1, 2002

By Dave Bulled

"This paper describes the design of a molybdenite flotation circuit using FLEET technology applied to kinetics data produced from laboratory flotation testing and distributed across the blocks of a mine plan using a geostatistical approach. The geostatistics assigned the statistical error that arises from having a limited amount of sample testing data to the kinetics estimates of each block of ore in the mine plan. The final design then included a quantification of risk based on the accuracy of the block flotation kinetics and model fitting. The design made reference to the extensive technical reports describing earlier metallurgical test work including some pilot plant work over thirty years ago.The testing involved conducting MinnovEX Flotation Tests (MFT) to determine the rougher flotation kinetics at the specified reagent conditions and rougher-cleaner tests to determine the change in kinetics resulting from regrind. These flotation kinetics were used within FLEET simulations to create a full-scale flowsheet and plant design including recommended equipment sizes and the forecast of operating results. Measurement of detailed flotation kinetics for a wide range of samples allowed the effect of size of grind and efficiency of classification (i.e. width of size distribution) to be predicted for the whole ore body.INTRODUCTIONAdanac Moly Corp. proposed to develop a 20,000 ton per day mill to produce high quality molybdenite concentrate from its Ruby Creek resource in north-western British Columbia. It is a relatively clean low grade porphyry-monzonite occurrence containing minimal quantities of other sulphides such as chalcopyrite, pyrite, sphalerite and galena. Molybdenite occurs mainly as coarse flakes with lesser amounts of finer sized values. An optimum plant design was required to can achieve the predicted operating results, at a known level of risk, while treating the variability inherent in the ore body."

Jan 1, 2007

By F. Vagnon, A. M. Ferrero, M. Pirulli, A. Segalini

"The topic of this study is the design of flexible fences to mitigate the risk due to debris flow phenomena by means of theoretical and experimental studies analysed in framework of the Rock Engineering System (RES).The study of the interaction between debris flows and flexible barriers is a major challenge since many different aspects concerning both the slope characteristics and the structure features are involved. Debris flows are often triggered by the mobilization of debris deposits that are generated by rock fall of adjacent rock walls. The rock debris deposits are difficult to be characterized by the physical and mechanical point of view due to their large variability in volume distribution and, moreover, when dealing with high mountains environment, they are also strongly affected by the presence of water at different phases (solid or liquid) depending on the temperature. The debris deposit turns unstable in relation with the geotechnical and environmental features of the areas and the instability evolution in high mountain steep valley is often a debris flow. Debris flow destructive potential is very high and protections are often needed to reduce the debris flow risk.This complexity determines the needs of comprehensive study able to include the global basin dynamic and the environmental conditions. An a-priori analysis of the influence of the different aspects can be useful to focus on the real driving aspects of the problem. Due to the complexity of the system the application of the RES methodology is recommended in order to clearly define the different assumptions needed, their influence on the simulations and a comparison among them.RES allows investigating triggering criteria, flow and depositional processes and interaction of flows with protection fences in a rational way. For this purpose, an interaction matrix is created: it summarizes the key geotechnical parameters that influence debris flows, their interactions and debris/engineering protection opera behaviour. The goal is to quantify the diagonal-off terms to assess the residual risk after barrier construction, using back analysis from real debris flow event and data collection from literature.This methodology is applied to analyze a debris flow event in the Italian Western Alps where numerical analyses are carried on to assess flow characteristics to be used for fence design through the application of an analytical and numerical model developed by the authors (Brighenti et al, 2013). In particular, the velocity and deposition heights have been estimated using the numerical code RASH3D (Pirulli, 2010), which is a single-phase model based on depth-averaged Saint Venant equations. Pathways of different lengths are studied to assess the input data necessary to design the protection barrier and to determine the associated risk. Simulation of the barrier at different places along the valley will also allow to determine optimal barrier location in terms of minimum residual risks by comparing the velocity, the volume and the energy of the debris after the barrier impact.The work will produce a possible design scheme to be adopted in this environment."

Jan 1, 2015

By T. S. Heah, S. T. Bishop

"The Lorraine property is located in the northern Intermontane Belt, north-central British Columbia. The property is underlain by intrusive rocks of the Duckling Creek Syenite Complex, an Early Jurassic, alkaline phase of the Late Triassic-Early Cretaceous Hogem Batholith. Two significant zones of copper-gold mineralization have been explored to date; the Main zone (historical Upper and Lower deposits) and the recently discovered Extension zone. A combined geological resource was estimated in 1975 for the Main zone deposits at 10 million tonnes averaging 0.7% Cu, and from 0.1 g/t Au to 0.34 g/t Au. The Extension zone, still at an exploration stage, is estimated to be considerably smaller than the Main zone.Copper-gold mineralization at Lorraine Mountain exhibits characteristics suggestive of both magmatic and hydrothermal origins, and may be related to orthomagmatic-hydrothermal fluid flow contemporaneous with magmatism and development of migmatitic fabrics. A depth greater than the depth of formation of classical porphyries may explain many of the variations displayed by mineralization at Lorraine. These characteristics distinguish Lorraine from other members of the alkalic suite of porphyry Cu-Au deposits in the Canadian Cordillera.IntroductionThe Lorraine prospect is located in the Omineca Mountains, at the headwaters of Duckling Creek, 56 km northwest of Germansen Landing (Fig. 1). The property is centred at latitude 55°55'N, longitude l25°26'W on NTS mapsheet 93N/14W. The area is typified by mountains of moderate relief, with elevations ranging from 1150 m in the valleys to peaks of 2000 m. Valleys are U-shaped and blanketed by glacial till which give way to steep, talus covered slopes and sharp ridges. Outcrop is generally limited to elevations above 1600 m (Fig. 2)."

Jan 1, 1995

By R H. Hallam

The Mattagami mining district encompasses a favour-?able belt of folded Keewatin-type rocks, with intercalated sediments and intrusives of late Keewatin age. This environment has been subjected to periods of basic and acidic intrusion. The orebodies lie within the fold structure and are associated with faulting and shearing of considerable magnitude. The No. 1 orebody contains 60 to 70 per cent massive sulphides. The minerals, in their order of abundance, are generally: pyrite, sphalerite, pyrrhotite, magnetite and chalcopyrite. It is evident that the broad geological structure of the district is amenable to ore deposition. There are two ? prominent linear expressions which comprise this broad structure. Both strike directions lie at the edge of, or are related to, a complex gabbro-anorthosite-peridotite zone. The fact that eleven of the ore zones discovered to date coincide with the broad structure indicates its over-all importance. It is estimated that the production from Mattagami Lake Mines will be in the range of 600 to 700 tons of zinc concentrate and 70 tons of copper concentrate per day.

Jan 1, 1964

By K. Awuah-Offei

Long-term production scheduling optimization has been a challenging issue for the mining industry because of the size and complexity of the problem. The current planning algorithms have limitations in terms of addressing the stochastic variables underlying the mine planning problem. In this paper, an intelligent agent-based mine planning framework based on reinforcement learning is introduced. The long-term mine planning is modelled as a dynamic decision network. The intelligent agent interacts with the block model by means of a stochastic simulation and employs a Q-learning algorithm to learn the sequence of pushbacks that maximizes the net present value of the mining operation. The intelligent open pit simulator (IOPS) was implemented with an object-oriented design in Java®. A comparative application case study was carried out to verify and validate the models. The proposed method was used in planning an iron ore deposit and the results were compared to the Milawa scheduler used in Whittle® software. The outcome of the study demonstrates that the intelligent agent framework provides a powerful basis for addressing real-size open pit mine planning problems.

Jan 1, 2008

By A. L. Mular, R. P. Edwards

"The operation of flotation circuits is often regarded as a combination of art and science. Expert systems offer a method by which the art may be captured and supervisory control of flotation circuits fully realized. Operating knowledge, the art, can be consistently employed to enhance regulatory control through the judicious application of an expert system. Using an expert system, a supervisory control routine was developed, and successfully implemented, in the copper flotation circuit at Brenda Mines Ltd.The expert system ran in conjunction with regulatory PI controllers on a personal computer. The expert system received assays via an on-stream analyzer and adjusted regulatory loop setpoints as required in real-time. From its commissioning in November 1989, to the end of the mine's operation in June, 1990, the ES was enthusiastically embraced by the operators and attained operating levels superior to those achievable manually.INTRODUCTIONExpert system (ES) software springs from research into artificial intelligence and is currently making significant inroads into business and industry. Proponents of ES's make many attractive claims; they are a means of capturing the knowledge of experts and they can be programmed by non-programmers. To date, most uses of this software in the process industry are in off-line applications, meaning that the applications are not tied directly to operating (on-line) environments. Moreover, in on-line environments, they are typically employed as advisors and only suggest changes. Human operators are required to close-the-loop to the process. Yet, ES's should be applicable to operating environments and, if the claims of the proponents are valid, should also be better than existing tools. As an intelligent controller an ES must use the knowledge of experienced operators, apply it directly to a process and do this without the intervention of human operators. On-line applications have been successfully employed in the closed loop control of SAG circuits (Herbst et al., 1989) and flotation circuits (Hales, 1989). Harris and Kosick (1988) detail an implementation for flotation control at the Polaris Mine."

Jan 1, 1991