Search Documents

By Kugan Kuganathan

Pastefill is being increasingly considered as a viable fill by underground mine operators. Mount Isa Mines (MIM) chose pastefill to fill their deep copper mine (>1200m deep) after considering various other types of fill in the feasibility studies. During the development of pastefill at Mount Isa (1996 - to date) the Mining research of MIM developed two cost effective, reliable and simple testing methods and used. With the help of these tests the behaviour of MIM pastefill was studied thoroughly and the most optimum mix design for pastefill was achieved. The MIM's pastefill plant, the largest in the world with a capacity of 450rph continuous pastefill production and unique in the process used, has been built (1998-1999) to produce pastefill to the specification developed through extensive testing program using the Flow Cone and Mini Slump cone Tests. As a consequence, it is now possible to select the optimum pastefill mix designs and predict their flow behaviour in bore holes or pipe lines using only these tests. MIM Blended Tailings Pastefill (BTPF) uses 3-3.55 Portland cement and the cost of pastefill is much lower than the total tailings pastefill, that could be produced from the same tailings.

Jan 1, 2001

By Deepak Malhotra, Linda K. Baltich

The minerals business is dynamic and competitive and the economic viability and/or profitability of any mineral company depends on maximizing recoveries while minimizing costs. In order to achieve this goal, it's essential that you have a clear understanding of the current status, including strengths and weaknesses, of the operation. The authors have developed a systematic approach to plant auditing which we have used successfully at several operations. The strategy presented in this paper can be tailored to fit a wide range of audits. This flexibility is a key feature of the approach.

Jan 1, 1989

By J. B. Zachwieja, P. E. Richardson, G. W. Walker

The electrochemical nature of mineral-collector interactions and their relationship to flotation have been investigated to develop a better understanding of the factors controlling sulfide mineral beneficiation. Spectrophotometric and electrochemical techniques have been used to examine the reactions occurring on electrodes composed of a bed of bornite particles in the presence and absence of potassium ethylxanthate (KEtX) at a pH of 9.2. The combined techniques showed that KEtX sequentially reacts with bornite by (1) a chemisorption reaction forming ca hydrophobic species between --0.4 and -0.2 V (SCE), (2) an electrochemical oxidation reaction forming cuprous xanthate between -- 0.2 V and the xanthate/ dixanthogen redox potential, and (3) oxidation to dixanthogen when the potential exceeds the reversible value for the xanthate/dixanthogen couple. The flotation response of bornite closely parallels spectrophotometric and electrochemical evidence for adsorption of KEtX by the reactions listed above. The degree of surface oxidation resulting from the formation of iron oxides on the surface dictates the flotation response as evidenced by the fact that oxidized bornite floats at more anodic potentials than unoxidized bornite. The strong correlations between the flotation response, degree of oxidation, and potential indicate that chemisorbed xanthate produces a hydrophobic layer sufficient to float unoxidized bornite, but cuprous xanthate and dixanthogen arc required to float oxidized bornite.

Jan 1, 1988

By D. W. Park

Artificial intelligence has emerged over the past two decades as an important problem-solving tool including techniques such as expert systems, neural networks, tree classification, fuzzy logic, genetic algorithms, etc. These methods can be employed to solve rock mechanics and ground control problems associated with the highly complex nature of rock masses and ground conditions. A current study at the University of Alabama focuses on development of an artificial intelligence system for real-time interpretation of the data collected from longwall monitoring stations, and development of an early warning system for prediction of undesired conditions and prevention of longwall-mine instability problems. This paper describes a ground control management system using UALSAS (The University of Alabama Longwall Stability Analysis System), which is being developed to facilitate maintenance, on-line analysis, prediction, and includes an early warning system. A detailed framework of computer algorithms and their applications are also discussed. The successful application of these artificial intelligence techniques will contribute to a safer and healthier mining environment.

Jan 1, 1995

By Gary F. Meenan

The industrial practice of froth flotation for upgrading the quality of fine coal is a relatively complex, expensive, and inefficient process when compared to the coarse- and intermediate-size coal cleaning circuits. The differences in cost and process efficiency can be attributed mainly to the lower equipment throughput, the added cost of flotation reagents, and the fact that flotation is typically a manually operated process susceptible to fluctuating feed characteristics that influence the overall operating efficiency. The emphasis in recent years has been on improving the flotation process efficiency and lowering the cost per clean ton. This has been achieved by larger flotation cell designs, implementation of flotation reagent evaluation programs, and improved process instrumentation and control. These developments are reviewed along with other current coal flotation practices.

Jan 1, 1999

By J. W. Merks

Spreadsheet software is an effective tool for simulating dynamic stochastic systems such as mineral processing plants. The two-product formula is used to show how to set up a Monte Carlo simulation in a spreadsheet template and how to calculate variances of dependent variables. The key to a reliable and realistic simulation model is to obtain unbiased estimates for all independent variables and variances in the set that defines the stochastic system. Monte Carlo simulation is also a powerful tool for optimizing sampling procedures, for applying sensitivity analysis and for designing and evaluating experiments.

Jan 1, 2000

By David B. Campbell

INTRODUCTION This submission was made to the Royal Commission of Inquiry into uranium mining in British Columbia. It discusses the stability of waste rock piles that might be developed in conjunction with uranium mining and the control of leachates that might be generated within these waste rock piles over the short and the long term. Short term refers to the interval during which a waste rock pile undergoes either continuous, or alternatively intermittent development, and in general can be considered to extend over the operating life of the mine. Long term refers to the postmining period, i.e. after the ore-body has been exhausted, and mining operations have ceased. Two major classes of waste material are generated in the course of mining and milling operations. Tails, which are the solid waste materials produced by the milling operations commonly exit the mill in slurry form and are stored in tailing retention ponds. Management and control of these slurried wastes are the subject of other papers to the Commission and this conference. In the course of the mining operations, it may be necessary to remove rock materials that are below cut-off grade. That is, ore concentrations within some of the mined rock may be lower than the concentrations which on the basis of economic considerations justifies milling the material. Rock which is removed in the course of the mining operations, but which is below cut-off grade is referred to as waste rock. This waste rock is commonly consigned to a specific area for disposal where it is placed in a mound or pile, i.e. a waste rock pile. Ore concentrations within the waste rock may vary from zero to just below cut-off grade. Since an increase in the market value of uranium could at some time in future result in reduction of the cut-off grade, it follows that waste rock containing ore concentrations slightly below today's cut-off grade could become ore grade rock at some time in future. For this reason, the mine operators may deem it advisable to separate waste rock according to grade, and to consign the waste rock to two or more individual waste rock piles. For purposes of clarity and simplicity, this submission treats the stability of waste rock piles and leachate control in a manner which assumes that these potential problems are isolated and separate from other aspects of the mining and milling operations. For many site specific cases, the problem of potential leachate control might be integrated with the tailings management aspect of the project, thereby realizing economic benefits while at the same time reducing the potential for adverse environmental impact.

Jan 1, 1980

By Michael H. Vallez

On September 13, 1973 MAPCO Inc. leased from Pocahontas Land Corporation an 18,000 acre tract of coal lands in Eastern Kentucky's Martin County (See Figure 1.). In July, 1974 initial run of mine coal shipments commenced and in April, 1976 shipments of a 12,500 BTU, less than 1% sulphur, washed coal was being shipped. Results of exploration demonstrated that large quantities of overburden would have to be moved. MAPCO decided to invest in large surface mining equipment to tackle the job.

Jan 1, 1981

By Donald S. Bubar

The Separation Rapids property, located approximately 60 km (37 miles) north of Kenora, Ontario, Canada, is host to one of the largest rare metal pegmatite deposits in the world. This pegmatite was named the Big Whopper by its discoverer, Dr. Fred Breaks of the Ontario Geological Survey, because of the exceptional length and width of the surface exposure. The property is readily accessible from a main, all-weather road via a network of secondary logging roads. The main line of the Canadian National Railway passes through the village of Redditt, just 40 km (24 miles) south of the property. The Big Whopper is one of the complex-type (petalite sub-type) class of rare metal pegmatites that are geochemically the most highly evolved in the spectrum of granitic pegmatites. Such deposits are economically important as resources for the rare metals lithium, tantalum, cesium, and rubidium. While complex-type pegmatites are found in many areas of the world, most are too small to be profitably mined, however, with an inferred resource in excess of 15 million metric tons (mt) (17 million short tons [st]), the Big Whopper is only the fourth deposit of its type in the world with the size required to be of major economic importance. The other three deposits, which are currently in production, are the Tanco mine in Manitoba, the Bikita mine in Zimbabwe, and the Greenbushes mine in Western Australia. The Big Whopper is situated in the eastern continuation of the Archean Bird River greenstone belt which also hosts the Tanco pegmatite, approximately 60 km (37 miles) to the west. The principal commodities identified in the portions of the Big Whopper pegmatite explored to date are petalite (LiAlSi4010) and rubidium-rich K-feldspar. These are industrial minerals with important applications in the glass and ceramics industries. The Big Whopper also contains substantial quantities of lepidolite, a lithium, rubidium-mica which is the principal ore mineral for rubidium metal. Tantalum and cesium occur in anomalous levels with the petalite mineralization, and the possibility of finding zones of high enrichment in these valuable high-tech metals elsewhere in the deposit is excellent. Such enriched zones are typical of pegmatites as highly evolved as the Big Whopper. To date, the geological mapping and diamond drilling work completed by Avalon have delineated the Big Whopper pegmatite system over a strike length exceeding 1.5 km (0.9 miles), over widths ranging from 10 m to 80 m (30 feet to 260 feet) in thickness, and to a vertical depth of close to 300 m (1,000 feet), where it remains open. The pegmatite system consists of a vertically oriented massive petalite pegmatite dyke striking 280 degrees that is flanked by amphibolites containing a swarm of narrower albite and petalite dykes which have all undergone intense deformation in a high strain zone resulting in folding and intense shearing. The Big Whopper exhibits a mineralogical zonation pattern which is complicated by the superimposed deformation that has both folded and stretched the deposit. Drilling has defined a dilute geological resource totaling 7.1 million mt (7.8 million st) grading 1.283 percent Li20, 0.346 percent Rb20, and .007 percent Ta205 over a strike length of 600 m (2,000 feet), and to a vertical depth of 250 m (820 feet), where it remains open. The petalite and rubidium-rich-K-feldspars contained in the Big Whopper all appear to be of superior quality. The grades are consistent with a petalite content averaging 25-30 percent and a rubidium-rich-K feldspar content averaging 15-20 percent. The remainder of the rock consists mainly of albite, several types of mica, and quartz. Accessory minerals include columbitetantalite, cassiterite, apatite, garnet, and gahnite. Market studies show that the deposit is located close enough to existing transportation infrastructure to access rapidly growing, major markets in the glass and ceramics industries, both in the northeastern U.S. and Europe. There are only three competing lithium minerals producers in the world with the Bikita mine in Zimbabwe being the only major producer of petalite. The other lithium deposits produce spodumene, a less desirable lithium mineral. Lakefield Research Limited has successfully designed a process to produce an ultra-pure petalite concentrate containing up to 4.65 percent lithium oxide, and as little as 0.014 percent iron oxide, low levels of soda and potash, and negligible amounts of other trace elements. With these specifications, Avalon will have an excellent quality product for glass-ceramics applications such as Corningware®, CERAN® stove tops, and other thermal shock-resistant products. Lakefield is presently designing the balance of the flowsheet to produce separate concentrates of rubidium-rich K-feldspar, albite, mica, tantalum, tin, and high-purity quartz. Initial test work indicates that the feldspar products will be of exceptional quality. The K-feldspar concentrates contain 11-12 percent K20 and 1 percent Rb20, while the sodaspar contains over 10 percent Na20, and both have very low iron (0.01 percent Fe203). A preliminary economic model shows that with an initial production rate of 90,000 mt (100,000 st) per year, expanding to 170,000 mt (190,000 st) per year over 5 years, the project is capable of generating over CDN $10 million per year in pre-tax cash flow by year 5 on a capital cost of approximately CDN $30 million.

Jan 1, 2001

By C. B. Rash, W. W. Fowler, Gill Montgomery

INTRODUCTION Fluorspar is the common term used for the mineral fluorite, which is naturally occurring calcium fluoride (CaF2). It is the principal source of fluorine, the most reactive of the chemical elements, from which a great variety of unique and strategically important chemicals are produced. Commercial deposits of fluorspar are widespread throughout the world, and its availability is essential to the industrial progress of all nations. GEOLOGY AND MINERALOGY In its pure form fluorspar, CaF2, contains 51.1% calcium and 48.9% fluorine, and has a sp gr of 3.18 and a hardness of 4. It is commonly a glassy mineral. colorless, white, or grayish. but may also be various shades of purple, pink, blue, green, yellow, or tan. It belongs to the cubic system mineralogically, and commonly crystallizes into cubic shapes in vugs and cavities, although most of it is massive, with interlocking crystals. Minable fluorspar deposits, in descending order of importance, occur as bedded limestone replacement deposits along fault zones; fissure-filling vein deposits; breccia filling in limestones, dolomites,

Jan 1, 1985

By Brain W. Lawrence

Sublevel stoping generally is a large-scale open stoping method. It sometimes is referred to as long-hole or blasthole stoping. This method usually is applied to strong ore bodies that require minimal support and are surrounded by strong country rock. The ore body should be fairly regular in shape and well-defined. The dip of the footwall normally should be sufficient to allow broken ore to gravitate freely, although the method has been adapted successfully to mine some of the flatter ore bodies. Sublevel stoping is not dependent on the width of the ore body. Widths of less than 6 m (20 ft), however, make the utilization of long-hole drilling techniques more difficult. The main criteria for sublevel open stoping are competent ore and stable host rock, regular ore boundaries, and a footwall dip that exceeds the angle of repose of broken ore. Basically, the method entails providing access to the ore body at various subintervals between the main haul- age levels in order to drill and blast the intervening ore. Stope drilling is carried out from drilling drifts on the sublevels, and the ore is blasted in slices towards an open face, which generally is vertical on the downholes and may be inclined towards the open face for the up holes. The blasted ore gravitates to the bottom of the stope and is collqcted through drawpoints. Fig. 1 illustrates a typical sublevel stoping mine. The method requires extensive ore body development with relatively high capital expenditures. How- ever, much of the development is in ore, and production costs are comparatively low. Productivity rates are in the 13.6 to 27.2 t (15 to 30 st)/manshift range. The drilling, blasting, and loading operations are performed independently, and equipment utilization is high. Large outputs can be obtained with few units and limited personnel. Dilution with waste rock may occur if ore boundaries are irregular or if caving occurs, but 100% of the ore within the stope usually is recovered. Pillar recovery sometimes is a problem. It is at this point that most of the waste dilution occurs. At the time data was gathered for this report (1973), the sublevel open stoping method was used by more than 21% of the metallic ore mines in the US (9 out of 42 mines producing 1088 + t/day (1200 + stpd) and provided more than 12% of the total metallic ore production from these mines. DEVELOPMENT Access to a sublevel open stope mine is gained by a slope or shaft that is normally sunk in the footwall of the ore zone, away from any possible effects from blasting or other production operations in the stoping process. Main level interval is selected and usually varies from 45.7 to 121.9 m (150 to 400 ft), depending on the vertical extent of the ore body, the number of drawpoints required to maintain output, and the eventual stope height. The main haulage drifts are located at the bottom of the stope horizon, either directly in ore or in the footwall, with access crosscuts at intervals. Raises are driven up through the ore to connect to the level above and provide access and ventilation to the sublevel drilling drifts, which are driven horizontally in the ore for the length of the stope. Depending on the drawpoint system to be used, the stope is either undercut by a raise-and-cone system, from a scram or slusher drift, or simply provided with a drilling drift from which access is made at intervals to the main haulage level. A raise driven either at the end or in the middle of the stope (if stope access is from both ends) provides a start for opening up a vertical slot across the stope to establish a free or open face for long-hole drilling and blasting. Sublevel stoping has proved to be a relatively safe and economical system.

Jan 1, 1982

By J. J. Komadina

Since the advent of resin-in-pulp (RIP) technology for treatment of uranium-bearing ores in the 1950?s to current carbon-in-pulp (CIP) processing of gold values, metallurgies and operators alike have wrestled with the problem of interstage screening. Early gold CIP plants such as at Homestake utilized externally mounted vibrating screens in conjunction with air lifts to effect carbon-pulp disengagement. Other operations world-wide have likewise adopted this technique with varying degrees of success. Subsequent development of in-tank launder screens both in South Africa and the United States have been of the air cleaned variety (EPAC)

Jan 1, 1988

By Robert L. Randall

The mineral industries have a continuing need for market forecasting and analysis for a variety of purposes, among which are organization and control of the sales effort, financial and corporate planning, and evaluation of the anticipated revenue from investment in proposed projects. For anything beyond the simplest extrapolation, some kind of economic reasoning or analysis is called for, which may or may not be incorporated into a formal or quantitative model. The principles described here apply to most of the commercially significant minerals and metals. The examples are drawn largely from the copper market, as it is the one with which I have the most practical experience. Practical forecasting is the process of analysis and synthesis from three categories of knowledge: facts and statistics, theory, and practical knowledge and informed opinion. Although many attempts have been made to build forecasts on only one of these three pillars, none has been successful as an operating procedure because of the lack of balance. There are no facts about the future-only about the present and the past. Facts can illuminate the future, but the future can be described only in terms of probability or more or less likely states of affairs-not as a continuous extension of the past and present. Likewise, theory, although useful as an organizing discipline and for providing a methodology of analysis, is inadequate for the complexities of reality. This is especially so when there is feedback between plans and forecasts to produce a new outcome and when there are significant elements of unpredictable free will, such as presidential jawboning or foreign policy considerations. Such elements are difficult to accommodate in neat theories, although they must certainly be reflected in the forecasts. Pure opinion frequently suffers from both a lack of-disciplined reasoning and the presence of biases or self-interest. While the future is unknown, it tends to cast its shadow before it; hence, forecasting is the art and science of positioning oneself so that the shadow becomes visible to the forecaster.

Jan 1, 1979

By D. E. Ranta, Stephen W. Thomas, Clancy J. Wendt

One of the most difficult tasks associated with placer mining is the sampling of the deposit. More placer projects have failed due to inaccurate assessment of the reserves than to any other reason. Within the realm of placers, those containing valuable minerals with a high unit value are more difficult to sample than those with larger bulk, lower unit value minerals. Some items to consider when sampling a placer deposit are as follows: A. A relatively large sample is needed for accurate valuation of the ground being tested. Placers are composed of many sizes of gravel which makes a representative sample difficult to obtain. B. When sampling placers for high unit value minerals such as gold, any error in mineral content of the sample will be magnified in the calculation of reserves. C. Values are usually erratically distributed within the gravel mass. Therefore, some placers with a more uniform value distribution may be adequately assessed with a minimum number of samples whereas a deposit with a highly erratic distribution of values may not be adequately sampled regardless of how many samples are taken. D. The investigation of a placer deposit should be made by, or be under the direction of, a person experienced in the art of placer sampling. F) During a sampling program, items that must be observed and noted in addition to the sample size and valuable mineral content should include boulder size and number, clay content, bedrock conditions, water, frozen ground, false bedrock, and any other physical characteristics that would affect mining of the deposit.

Jan 1, 1992

By R. S. Shoemaker

It is said that there is nothing new under the sun and this is certainly true about refractory gold Ores. Ores which have been refractory to treatment have been the bane of the Metallurgist's life ever since gold has been recognized as a metal valued for its indestructability, malleability and beauty. Webster's Dictionary defines "Refractory" as; "Resisting control or authority, stubborn, resistant to treatment or cure, unresponsive to stimuli, difficult to corrode or draw out," And gives as a synonym the word "unruly." I don't think any metallurgist would disagree with these definitions as they quite accurately describe refractory gold ores as we know them and I wouldn't be a bit surprised if many of you in this room today had applied far stonger language to your own refractory ores in the course of your metallurgical test programs. Pliny the elder put it very well when he wrote in 79 A.D., Gold is the first of man's follies, Silver the second." Perhaps Pliny was speaking of "unassayable gold" or "incipient gold." These are truly refrac tory ores which probably existed before Pliny's time and although many people have claimed to have solved their metallurgy the untold wealth that lies within them has never been bankable. On the other hand, Virgil is his Aenead, could very well have been speaking of refractory ores when he wrote; "Quid non mortalia pectora cogis, auri sacra fames!" which translates as, Accursed thirst for gold! What doest thou not compel mortals to do. Certainly, gold was extremely rare at that time because it had to be recovered by much labor from the then refractory ores and their refractoriness caused the early metallurgists (or alchemists if you will) to seek an easier way to produce it; thus the long and fruitless search for the "philosopher 's stone" which would aid in turn-

Jan 1, 1990

An upgrade to Vale?s existing Clarabelle dual-pinion, semiautogenous (SAG) mill was completed by the power and automation technology company, ABB. The mill is used to reduce the size of the ore at the company?s nickel and copper mines in the Sudbury area. The project was completed on time and on budget with an excellent safety record. Located in Ontario about 400 km (250 miles) north of Toronto, Vale?s Sudbury operations are primarily underground nickel sulfide mines, with integrated mining, milling, smelting and refining operations. One of the characteristics of Clarabelle Mill is to accept nickel-copper ore from various sources and companies and to produce a nickel-copper concentrate and copper concentrate for further processing. The size of the ore is further reduced in SAG mills, rod mills and ball mills, where all milling is done wet.

Mar 1, 2014

By Osvaldo A. Bascur, J. Pat Kennedy

Large metallurgical complexes have more data than a small city and much of it is "real time," changing rapidly as internal and external conditions evolve. As corporations buy and sell assets and reengineer their staff, the structures and tools that were used to understand these data are lost. The lack of an integrated approach makes the profitability of metallurgical plants lag behind that of other industries despite technical advances that have occurred in recent years. This results in significant deterioration in the ability to maintain a "smart” operation. The solution is to use the sophisticated new tools built for the Internet so as to leverage information in a familiar browser format that people can use and understand. Combining real-time infrastructure with portal technology is becoming an accepted methodology. This paper introduces real-time portals and team sites to support decision making by work- groups which require both real-time and documentary information from manufacturing assets, vendors, head office and the Internet. Workers get insight from the overall process effectiveness index based on key variables such as losses due to plant availability, performance efficiency and rate of quality. Real-time portals and smart clients are part of any real-time performance management (RtPM) solution. Three case studies in large metallurgical complexes are presented. PLANTS ARE YOUR STRATEGIC HANDICAP Information technology is arguably one of the fastest moving technologies used by the mining/metallurgical industry. However, the business model for managing that

Jan 1, 2004

By Thomas M. Larimer

Prior to April, 1961, the coal charge to Clairtonts coke oven batteries was 3" x 0" mine waslied coal having approximately 6.0% to 7.0% moisture. Even at this size and moisture content, the arching, rat-holing and hang-up was excessive and the effective live storage of coal was only 60% to 70% of the rated capacity of the coal bunkers, depending upon the moisture of the coal. In addition, many production delays were encountered in the time and methods required to fill the larry cars. The need for an improved metallurgical coke led to the introduction of pulverized coal to a portion of the plant in 196 1. When the full potential of the pulverizing equipment was realized, we were able to serve seven batteries in our second unit of batteries with pulverized coal through two coal bunkers. The discharge hoppers of these bunkers were equipped with air vibrators to assist the flow of pulverized coal into the larry cars. The vibrators were helpful in this respect but required considerable maintenance and caused damage to the bin plates.

Jan 1, 1968

By Neal Cochran

I consider it a distinct privilege to be afforded the opportunity of discussing the program of the Office of Coal Research. The principal activities I will be discussing are concerned with projects that fall in the utilization area. I will, to be sure, touch on other activities of the Office of Coal Research but am not equipped to discuss these activities in great detail. Research activities of the Office of Coal Research are conducted on a contract basis with commercial, academic, and non-profit institutions. We do not operate any laboratories of our own but we do maintain a very small staff for technical and administrative supervision of our contractors. We have consistently attempted to interest outstanding, creative, and practical research organizations in our program. We are constantly seeking the imaginative new idea that will materially reduce the cost of mining coal, transporting it, or will process it into a more convenient and useable form. In assessing proposals and in developing our program, we search for the improvements that will develop large-scale new uses for coal or coal products or those that support these new uses.

Jan 1, 1963

By James A. Finch

The Gaudin lecture is devoted to fundamentals. Given my involvement with column flotation over the last 20 years, this is my topic. I have tried to identify spinoff basic research and provide a personal list of favorites. Two are selected for discussion here. They are single bubble rise and bubble surface area flux. The rationale for such work, as well as some of the findings, are reviewed. Support for fundamental work seems as volatile as the economy (and the link is probably not coincidental). I hope some utility to the minerals industry is evident in the text. Column flotation in its current form originated in the early 1960s. The method was invented by Boutin, patented by Boutin and Tremblay and pursued doggedly by Wheeler. Their aim was to overcome the problem of mechanical entrainment of fine particles in the water carried by the bubbles, which they were convinced contributed to lowering flotation concentrate grades (Wheeler, 1988; Finch and Dobby, 1990). The solution was a combination of columnar geometry, the addition of wash water into a deep froth and the maintenance of an excess of water to the tailings, as compared to the feed (a so-called positive bias). Success at Les Mines Gaspe, Quebec, on cleaning Mo in the early 1980s signaled the start of industrial acceptance, making the column one of the flotation machines of choice today. Over the next 20 years, column applications mushroomed to cover virtually all mineral commodities, as well as nonmineral applications such as soil decontamination, deoiling of water and deinking of recycled paper. In the wake, research groups are continuing to probe the fundamentals, conferences and workshops have been transferring the technology and specialist companies have emerged to service the demand. Innovations were triggered as problems arose (for instance, in gas injection systems) and novel flotation cells, much in vogue in the 1990s, had the door opened. Mechanical cell manufacturers have not been idle - there is nothing like competition to provoke a fresh look at an established business. Flotation-cell technology has probably not seen so much activity since the patent rush in the early days of flotation, which occurred around the turn of the century. For an industry often considered reluctant to risk innovation, this is not a bad record. A good idea will find a place, but patience (not always an innovators strong suit) and careful preparation are required. Two books (Finch and Dobby, 1990; Rubenstein, 1996), three symposia (Sastry, 1988; Agar and Huls, 1991; Gomez and Finch, 1996), and a series of reviews (Dobby and Finch, 1991; Finch and Dobby, 1991; Finch, 1995; Finch et al., 1995) document the period. Like many good ideas, column flotation has been overexposed, and there have been disappointments. In a recent example, columns performing scavenging duty were replaced by mechanical cells (Chong et al., 1998). The further applications stray from the original - to

Jan 1, 1999