Today’s Technology: Canadian Mineral Processors: Latest Techniques Separate Good from Gangue
By all measures the 32nd annual meeting of the Canadian Mineral Processors was a success. The presentations were informative and thought-provoking; they covered operating experiences, grinding, flotation, new projects, expansions, and process control. About 350 delegates came from Canada, the United States and Europe to meet in Ottawa last January.
Members always enjoy operations papers such as the one about the Pierina deposit in Peru. This deposit was acquired by Barrick Gold in August 1996 when it purchased Arequipa Resources Ltd. Twenty-seven months later, on November 25, 1998, the first dor bar was poured. This is Barrick’s first greenfield project outside North America. The $316-million mine and heap leach and recovery facilities were completed on time and under budget.
The Pierina deposit is hosted within the remains of a volcano containing porous permeable acidic pumice tuff and andesite. When acidic hydrothermal solutions carrying gold and silver later came in contact with these rock types, two alteration zones were created. Where the solution came in contact with the pumice tuff, a vuggy silica was deposited, and this zone contains most of the mineralization. Where the rock was less reactive or the solution weaker, quartz alunite was formed, which is considered waste by virtue of its only-minor mineralization.
Metallurgical testing revealed that the gold occurs in a fine (< 0.1 micron) free state and leaches readily. Silver is found mainly as acanthite with only minor amounts of native silver. There is some mercury associated with the acanthite, and the mercury is thought to reduce the leachability of the silver. Further tests were conducted to determine the economics of heap leaching and carbon-in-leach recovery. The lower capital and operating costs of heap leaching, despite reduced metal recoveries, made it the method of choice.
The heap leach operation treats 19,500 tons per day, and may be expanded to 27,000 tons. Ore is crushed until it reaches a nominal -11/2-inch size. The crushed ore is hauled to the leach pad, which is raised in a series of 30-foot lifts. The Phase I leach area is sufficient for two years’ production, and the ultimate pad size will be 120 million tons.
The heap is leached with barren solution containing 0.5 gram per litre sodium cyanide. It is applied at a rate of 0.04 US gal/min per square foot per day. Pregnant solution is collected behind a waste rock dam lined with clay and two layers of geosynthetic liner. Full leak detection technology is employed. Seepage and groundwater are collected by an underdrain system and discharged to a collection sump. If it meets regulations, it is discharged as normal drainage. Should pond leakage contaminate this water, it would be pumped back to the pad or to the gold recovery plant.
Pregnant solution is treated in a Merrill-Crowe plant to recover gold and silver. The solution is clarified, stored and pumped to the top of a Crowe deaeration tower. When the solution is drawn off from the Crowe tower, its oxygen content has been reduced to 1 ppm from 5 ppm. The next step is to add zinc dust, cyanide and lead nitrate (if required) at the pump and direct the solution to plate filter presses in the refinery. Filtercake, containing gold, silver, mercury and other metals, is treated in a retort to vapourize mercury, which is recovered for sale. The retorted material is melted with flux in an induction furnace and poured into 1,000-oz dor bars.
Barren solution is collected and treated with cyanide, lime and antiscalant before being pumped back to the heap. As necessary, a portion of the barren solution can be bled to an alkali chlorination circuit to reduce cyanide levels. Provision has been made to add arsenic- and mercury-removal to this circuit.
Barrick has no reservations declaring the Pierina project a success. Performance thus far has been as predicted: 100,000 ounces of gold monthly at less than US$50 per ounce cash costs.
ON-SITE DAILY MINERALOGY
Small and medium-sized processing plants, especially those in remote locations, often lack timely access to mineralogical testing. That problem was solved at the Nchanga Division of Zambia Consolidated Copper Mines (ZCCM) with the implementation of an on-site daily mineralogy (Sidmin) program. The aim is to give operators this information every day, particularly when feed variations are significant. The program is built for speed and economy, providing meaningful results for practical purposes. It is intended as a supplement to, not a replacement for, chemical analysis.
The Sidmin program begins with the collection and drying of samples. The powder is tested without screening, briquetting and polishing because of the time needed for those processes. Samples are deslimed by stirring them in water, allowing the solids to settle, and pouring off the supernatant suspensions. The wet material is transferred to a sample plate for microscopic inspection. A stereomicroscope is used with coloured and opaque minerals or a polarizing microscope with colourless minerals. Chemical reagents are used for definitive testing on grains in case of doubt. Staff is trained in identification and counting mineral grains of interest. The volume of a given mineral is weighted by visual approximation, which speeds up the counting process but is difficult to master. Finally, daily reports are created and circulated.
ZCCM reports that its Sidmin program has been successfully used since November 1997. Originally, the reports were distributed only to plant superintendents, but demand grew. They are now given to the manager of metallurgical processes, the chief geologist, assistant chief geologists, and the mine managers. This sharing of information has led to increased communications within the processing department and between that department and others. The mineralogical information also helps management tackle problems with confidence.
As soon as Sidmin reports became available, operators noted persistent free sulphide particle losses in the floatable size range. Within two weeks the problem was remedied, and sulphide recovery increased more than 1%. Sidmin information also revealed that leachable oxide minerals were reporting to the tailings leach plant. That, too, was corrected, and leach efficiency has steadily increased.
EFFECT OF SHOTCRETE ON FLOTATION
Underground mines use shotcrete for ground support or concrete in backfill. That’s fine unless the cement materials reduce metal recovery in the flotation circuit, and there is ample evidence that this is the case. The bright minds at Laurentian University and Falconbridge’s Strathcona mill have studied the problem to learn why the contamination creates the problem, and also to find a means of removing it or reducing the effect. Other mining companies have reported lowered nickel recoveries due to cement contamination as low as 1%, but the precise reasons for this reaction were not understood until now.
The researchers reported having some difficulty in reproducing test results but could identify several trends. They believe that lime in the shotcrete is responsible for lowering nickel recovery as lime may hydrate and coat the mineral particles, elevating the pH above the optimum processing level. The younger the shotcrete the more it affects recovery. They note that adding sulphuric acid during flotation to control pH improves recovery.
Work should continue, they recommend, on adding acid to the roughers in plant-scale studies. More work should be done into the different additives in shotcrete, particularly accelerators. Various methods of mixing should be studied to determine what if any effect this has on recovery.
TAKE THE HARD WORK OUT OF BOLT REMOVAL
Changing the liner in a grinding mill is hard work and can ring up substantial costs. Removing the bolts that secure the liner has been identified as the most time-consuming and physically strenuous element of the procedure. Manual bolt removal requires first using a pneumatic bolt removal tool to unscrew the nut and then an impact hammer to push the bolt into the mill. Workers frequently complain of backaches, sore upper body joints and general fatigue.
Having studied the bolt-removal problem thoroughly, the Centre de recherche industrielle du Quebec (CRIQ) created its unique mill liner bolt-removal system (patent pending) to take the hard work out of the job. This is a fully pneumatic tool that is easy to operate and transport. It combines a non-impact device to unscrew the nuts, an impact hammer to eject the bolts, and a nut recovery mechanism. All this is mounted on a fixed base and operated by remote control. The tool that unscrews the nuts also replaces them, tightening each to a consistent, specified torque. The operator is left to position the equipment at each bolt and use the remote control to screw or unscrew the nuts and eject the bolts.
Already in use for several years, the CRIQ mill liner bolt-removal system has increased productivity and safety. It cuts the time it takes to change liners in any type of mill by about 15%. It works on the feed head, the discharge head and the mill shell. It is easy to operate and eliminates the risk of injury to workers.
GETTING A HANDLE ON CMC
Carboxy methyl cellulose (CMC) is used at the Thompson mill in Manitoba to improve pentlandite recovery. The sulphides in the ore are mostly in peridotite altered to serpentinite. Recovery of nickel-bearing pentlandite is adversely affected when the pentlandite grains are coated with serpentine slimes. Using CMC depresses the slimes and allows better flotation of pentlandite. However, adding CMC in a typical 1% to 3% solution requires a handling system so large as to negate any metallurgical gains in the circuit; therefore, an economical and effective method of dry CMC addition has been developed by Inco Technical Services.
Testing was conducted to assess how and where CMC addition is most effective, and also to determine optimum conditioning time. Similar results were noted when the reagent was added either as a 1% solution or dry. In tests where CMC was added in the grinding circuit, efficiency was reduced somewhat compared with adding it after grinding. When dry powder was added, efficiency went up with increased conditioning, but the greatest gains were seen in the first two minutes.
The next step was to design and build a system to add dry CMC into the ball mill discharge. The powder is purchased in one-tonne sacks and suspended from a steel frame. Using a vibratory feeder and a Teflon-coated screw meter, the CMC is funnelled into the circuit below the mill discharge trommel. The slurry is then cycloned, and about 40% of the water (and the CMC) is recirculated to the grinding mills. The remainder flows to the flotation cells and has an estimated residence time in the pumps and piping of about three minutes, adequate for conditioning.
ON-LINE VISUALIZATION OF FLOTATION
Understanding flotation is the Holy Grail of mineral processing. To that end, a team from the University of Alberta, with support from Suncor and Syncrude, has created a novel on-line visualization system. The technique allows for the observation of the loading of solids on bubbles.
The size and number of bubbles are important parameters when modelling flotation, and recently bubble surface area flux has been shown to have an even better correlation with the flotation rate constant. But the application of bubble surface area flux depends on how accurately it can be estimated.
In the lab a one-litre Denver cell was fitted with a glass sampling probe. The probe can be moved around the cell to sample bubbles from various locations. The bubbles can be examined in real time with closed-circuit TV or later using a video recorder. Before each test the probe is immersed in the flotation cell and allowed to fill to the desired level. The slurry is conditioned, and a predetermined amount of air introduced. This allows the viewing of bubbles, particles and bubble-particle aggregates under natural conditions as they enter the probe.
The team observed the settling of particle-bubble aggregates directly, thought to be a first, providing direct experimental evidence that overloading air bubbles may actually reduce recovery. This occurs when the buoyant force of the bubbles is less than the gravity force of the particles attached to them.
Three distinct bubble-particle attachment patterns were identified. A single particle may be attached to a single bubble. A few particles may attach themselves to a single bubble. Or a group of particles may be attached to a group of bubbles, forming a bubble-particle lump.
The average bubble size generated in the solution could be increased under two conditions that were tested. Hydrophobic coal was added to an MIBC solution, and silica was introduced to an amine solution.
MORE ON BUBBLE SURFACE AREA FLUX
Much has been debated in recent years about bubble surface area flux being the best, or only, variable that explains the flotation rate constant. Work done at the Helsinki University of Technology, however, indicates that bubble area flux by itself is not an adequate dimensioning tool. Studies uncovered that flotation results depend on particle size distribution and other traditional cell variables as well as bubble area flux.
Researchers developed a method, based on a set of equations, for obtaining the true flotation rate constant in the pulp. They determined that when a very thin froth layer is present, recovery appears to have a semi-linear response to bubble surface area flux. Thicker froths do not give such a linear response. They next factored in particle size, and determined that it affects recovery much more than bubble surface area flux.
They concluded that fine ores need and tolerate high superficial gas rates and high bubble surface area fluxes. Coarse ores reach a maximum point, and a very high superficial gas rate and a high bubble surface area flux will not improve results. Good results depend on matching the hydrodynamic conditions to the task at hand and being able to adjust the froth properties and transport.
The texts of all the papers presented at the CMP meeting are available in the Proceedings. Contact Angela Putz in Ottawa by phone (613-992-8895), fax (613-947-1200) or e-mail (aputz@nrcan.gc.ca).
Comments