The flotation column is arguably the most important piece of mineral processing equipment developed in Canada in the past decade. But its acceptance has until now been thwarted by the lack of a sound methodology for scaling up the technology from the pilot plant scale to the production scale. Research by Prof Glenn Dobby, of the University of Toronto, and Prof James Finch, of McGill University in Montreal, is beginning to provide that methodology. The flotation column is particularly attractive to Canada’s remote mining operations. Compared to conventional flotation cells, columns save both space on the mill floor and electrical power consumption. They even produce a higher-grade concentrate, which could result in lower transportation costs from remote mine sites to smelters.
A joint effort between Dobby, at U of T’s Dept. of Metallurgy and Materials Science, and Finch, at McGill’s Dept. of Mining and Metallurgical Engineering, has expanded applications of flotation columns by providing sound technical analysis of design and operating principles. They are developing a scale-up model for flotation columns which will enable mill operators to implement, with confidence, flotation columns on an industrial scale. This scale-up model, for example, has cleared the way for effective column cell utilization in the copper-cleaning circuit at Magma Copper Co. in San Manuel, Ariz.
The research is supported by three $63,000-per-annum Natural Sciences and En gineering Research Council co-operative research and development grants from the federal government. Cominco Ltd. is a collaborating industrial partner.
The flotation column concept was invented and developed in Canada in the early 1960s. Since then plant and laboratory columns have been marketed commercially by Don Wheeler of Column Flotation Co. of Canada. In 1980-81 at Mines Gaspe, in Quebec, two flotation columns replaced 13 conventional cleaners in a molybdenum-upgrading circuit, with superior results. Since then many mining companies throughout Canada (including Placer Dome, Inco Ltd. and Cominco Ltd.) have constructed and installed flotation columns, either as full-scale units or, more commonly, as pilot-sized units. An ambitious program is underway at Gibraltar Mines in British Columbia. In response to successful plant trials, three columns, 2.2 m in diameter, have been installed to treat all the feed to the copper- cleaning circuit there.
There is now strong interest in flotation column applications worldwide, in the U.S., Chile, Peru and Australia, with Canada developing most of the technology.
The flotation column differs dramatically from conventional mechanical flotation machines, both in design and operating philosophy. Industrial flotation columns are typically 13 m high and 0.3-2.2 m in diameter (either circular or square). Three zones are identified: collection, cleaning and froth. Feed slurry enters the collection zone 2-3 m below the top of the flotation column and descends against a countercurrent flow of rising bubbles generated by a gas sparger in the base of the column. After collision, the hydrophobic (water-hating) mineral particles will adhere to the bubbles and the bubble-particle aggregate will rise into the cleaning zone. Hydrophillic (water-loving), non- flotable material is removed from the base of the flotation column as tailings.
The cleaning zone incorporates a unique feature: water is added from an array of perforated pipes usually located below the concentrate discharge lip. This wash water drains downward, generating a packed bubble bed. The packed bubble bed is very efficient at suppressing gangue entrainment to the concentrate. The downward velocity of the wash water is regulated by the difference between the tailings flow rate and the feed flow rate (or bias). The component of the wash water which flows downward (to make up the bias) is sufficient to prevent bubbles from coalescing in the cleaning zone. The withdrawal of tailings from the bottom of the flotation column is usually controlled at a rate slightly greater than the feed flow rate (called a positive bias). A conventional froth zone, about 0.05 m thick, forms above the wash water addition point in the cleaning zone. The purpose of the froth zone is to transport particles to the launder.
The interface between the collection zone and cleaning zone is distinct and is maintained at a set level either by control of the wash water rate or the tailings rate. The distinction between the cleaning zone and froth zone is not always obvious; there could be a gradual transition from a packed bubble bed to a froth. The concept of the two zones, however, recognizes that there exist two significantly different functions: cleaning and concentrate removal. The combined cleaning and froth zones will be referred to hereafter as the cleaning zone. Typical superficial velocities (flow rate per cross-sectional area of flotation column) are: gas, 1-3 cm per sec; slurry in recovery zone, 0.5-2 cm per sec (down); and water in cleaning zone (the bias), 0.10-0.50 cm per sec (down).
Dobby and Finch have been developing a soundly based methodology for scale-up of flotation columns from laboratory data to the plant. The approach for modelling or scale-up is to consider the flotation column as consisting of two regimes, the collection and cleaning zones, and to develop an understanding of the mixing and the kinetics of each regime separately. Then the models of the two regimes can be combined to yield an overall column model.
The collection process in a flotation column can be studied by considering a gas bubble rising through a downward flowing slurry. Recovery in the collection zone is determined by three factors: the rate constant, the mean residence time of the particles in the collection zone, and the mixing conditions of the collection zone. One extreme of mixing is plug flow, where the residence time of all elements of the fluid and all mineral particles are the same. The other extreme is complete mixing. Flow conditions in a laboratory flotation column approach plug flow. The scale-up of the collection zone from laboratory data to plant has been modelled by Dobby and Finch. The model has three significant parameters; mean residence time, collection rate constant and the dimensionless vessel- dispersion number which describes the degree of mixing. Techniques have been developed to estimate these parameters.
The scale-up model of the cleaning zone of the flotation column is less developed than that of the collection zone. Dobby and Finch have been studying the problem, but a significant and complex difficulty remains, namely an incomplete understanding of the operating variables upon cleaning action in the water-drained froth. These variables include the bias, gas flow rate, bubble size, solids loading of the bubbles and the cleaning zone dimensions. Their work on the scale-up model of the cleaning zone is in progress. The model consists of a simple mathematical equation which relates recoveries in the cleaning zone and collection zone to total recoveries. Since both total recovery and recoveries in the collection zone are known, this relationship can be used to calculate the recovery of the cleaning zone. With further testing, they hope to develop a more detailed model for scale-up of the cleaning zone. So far, the study has revealed that there is a carrying capacity limit in the froth zone which must be considered in the scale-up model. This should lead to a more reliable overall scale-up model of flotation columns from laboratory data to the plant.
The flotation column scale-up model developed by Dobby and Finch provided a method that was useful in designing the flotation column installation at Magma Copper Co. In April, 1986, flotation columns replaced the conventional copper flotation cleaners at San Manuel. Two single-stage copper flotation columns replaced 36 flotation machines, 28 on the first cleaner and eight on the final cleaner.
Magma Copper Co.’s San Manuel Division is 72 km northeast of Tucson, Ariz. The orebody, a disseminated porphyry copper
deposit, is mined by deep-level block caving. Primary copper mineralization is chalcopyrite with trace amounts of chalcocite, bornite and covellite. Molybdenum occurs on fracture surfaces or with quartz veinlets. Pyrite is the major sulphide gangue mineral.
The flotation column scale-up model was used to determine the size of the copper cleaning flotation columns at San Manuel. The basic parameters used in sizing the cells were established on the flow rates and performance of the conventional 2-stage copper flotation cleaning circuit. Using these parameters, the calculations indicated that the desired results could be achieved if two flotation columns replaced the 36 flotation machines. One column was 1.5×12.1 m; the other, 1.8×12.1 m. The diameters were 1.5 and 1.8m.
After deciding upon the size of the flotation columns, an operating strategy was established. A test program involved evaluating the flotation columns over a wide range of operating conditions. The major areas investigate d were the effects of varied pulp level, wash water and air flow rate. The optimum range of each variable was established.
The installation of the single-stage flotation columns that replaced a 2-stage conventional copper flotation cleaner circuit at San Manuel produced similar metallurgical results: a concentrate grade of 29.8% copper with a sulphide copper recovery of 90.4% for the flotation column cleaner circuit as compared to a 30% copper concentrate grade and a sulphide copper recovery of 90.1% for the conventional flotation cleaner circuit.
The flotation column scale-up model developed by Dobby and Finch is one of the specific objectives in their global objective of developing a better understanding of the emerging technologies related to flotation columns. Other objectives include: development of a flotation column circuit simulator; assessment and improvement of existing bubble generation techniques; development of level and bias sensors; selection of appropriate stabilizing control and performance control strategies; and demonstration of advantages gained by concurrent flotation column operation.
Be the first to comment on "RESEARCH Miracle in the mill"