Sorting out low-grade nickel ores
Researchers in Canada have developed a sensor that has the potential to make low-grade nickel ore extraction more economically viable.
The high-frequency electromagnetic sorting system (HFEMS) is aimed at sorting the ores at the pre-milling stage, and has been developed by BC Research Ltd, in collaboration with the University of British Columbia and Vale INCO Ltd. A start-up company, MineSense Technologies, also based in British Columbia, Canada, is commercialising the technology.
The system is a variant of the basic metal detector and consists of two coils, a sensor and reference. Both are fed with a low-power, high frequency alternating current that generates an oscillating electromagnetic field above the coil. When a magnetic or conductive object comes within range of the sensor coil, it alters the coil’s magnetic field and electrical properties in proportion to the quantity of conductive material present.
The reference coil is kept isolated from the object, and the difference between the sensor and reference fields – the sensor response – is measured using an electrical bridge network, amplified and digitally analysed. This response is then correlated to a material type by comparing the anaylsed response to a database of readings from samples of tested content and grade.
The device is said to be sensitive to nickel and copper at accuracies of +0.1% and +0.2%, respectively.
Currently the industry tends to rely on processes such as dense-media separation (DMS) for ore pre-concentration, but sorting systems generally offer lower capital and operating costs, says Dr Andrew Bamber, CEO of MineSense Technologies.
He outlines, ‘Our database of potential ore types has been built up over three years of laboratory testing with the sensor and one year of commercial operations. Over 40 samples have been tested, representing 15 different types of nickel-copper ore, including Thompson Nickel belt, Sudbury Igneous Complex ores, Voiseys Bays, Raglan Nickel belt and others. Each sample is initially tested with the sensor, and analysed using the half-rock assay method to establish the sorting potential’.
He notes, ‘Total [operating] costs for DMS can be up to US$3.50/t at 5,000t/day. With our system, for a similar capacity plant, typical costs are US$1.75- 2.50/t. Capital costs for each process would be comparable, although the cost of multiple media circuits for a DMS process can become prohibitive at capacities above about 12,000t/day.
‘We have conducted several [industrial] studies into DMS separation and sorting, with benchmarks for both operating and capital costs coming from [these]’.
Professor Jan Cilliers of the Department of Earth Science and Engineering, Royal School of Mines, Imperial College London, broadly welcomes the technology. ‘The industry is always looking at either grinding material to a lesser degree or grinding less material. Anything that helps with that is a good thing. This sensor system, therefore, looks to me to be a useful tool, albeit as yet another tool in the toolbox.’
Capacity is a key issue though. ‘The real challenge is to make the technology work on an industrial scale, at thousands of tonnes per hour,’ Cilliers says. Bamber claims, however, that the system can achieve the capacities required, quoting operating frequencies of up to 1.2MHz and 4,000 samples per second overall.
He notes the key to achieving the capacities required is to integrate HFEMS sensing technology with conventional material handling equipment found in the mine, such as shovels, belts conveyors and vibrating feeders.
The team is currently conducting service tests of its MK II HFEMS sensor for Xstrata Nickel and Canadian Arrow Mines.