Near Infrared Reflectance (NIR) Ore-Sorting of German Kupferschiefer deposits into differing ore lithologies technical paper, AW1 are testing X-Ray sorting.
XXVI INTERNATIONAL MINERAL PROCESSING CONGRESS (IMPC) 2012 PROCEEDINGS / NEW DELHI, INDIA / 24 - 28 SEPTEMBER 2012
The German Kupferschiefer
Kupferschiefer deposits experience divergence in the elemental and mineral composition its basic build-up remains the same. It consists of three lithological layers: Sandstone, Dolomite and Kupferschiefer. The main ore bearing minerals are Chalcocite, Bornite, Covellite, Chalcopyrite, Tennantite, Galena, Sphalerite, Pyrite and Marcasite. The waste rock is composed of Dolomite, Quartz, Feldspars, clay minerals and organic substances, mostly organic carbon (Siemroth, 1999)). In general the surface color of the sandstone is gray, the Kupferschiefer is black and the dolomite has a more brownish-grey color. Each lithology´s color may vary from light to dark depending on the Corg content, due to the bitumen (Clement and Hammami, 1980).
Mineral processing of Kupferschiefer ore
The conventional mineral processing of the middle European Kupferschiefer up to copper grades of 27 % Cu, as daily conducted, for example by KGHM in Gliwice, Poland, includes coarse to fine grinding followed by flotation (Weiss, 1985). All copper bearing lithologies are mined and processed collectively. Several investigations, however, prove that a separate flotation of the sandstone ore leads to a better recovery and higher copper grades
in the concentrate. Because of its relatively coarse structure and simple mineral composition, the sandstone, being a siliceous ore, lends itself to flotation (Clement and Hammami, 1980). The difficulties of floating the Kupferschiefer arise from the following characteristics: the sulphides are either finely disseminated as individual grains or else are finely intergrown. A fine grinding down to -70 or even -40 µm is required to achieve sufficient
liberation (UVR FIA, 2011). Secondly the carbonaceous bitumen content in the dark shale ore increases the flotation collector consumption and decreases the Cu recovery in the concentrate (Hammami, 1980). A third aspect is that the dolomite´s alkaline host rock, MgCaCO3, causes a higher consumption of acidic pH regulator (Riedel, 2004).
Implementation of Sensor Based Sorting
With the help of sensor based sorting the mineral processing of the Kupferschiefer deposits can be optimized. The sandstone is to be segregated from the coarse fraction of the Run of Mine (ROM) and floated separately.
Studies of the Instytut Metali Niezelaznych (Institute of Non-Ferrous Metals) from 2011 confirm, that a separate flotation of the Spremberg sandstone lithology adjusted specifically to its properties is able to achieve a Cu grade of 50 % at a Cu recovery of 93 % compared to 32 % Cu grade and 85 % recovery with conventional flotation.
THE NIR SORTING
Compared to other sensor based sorting techniques, the NIR sorting alone shows promising results to separate the sandstone from the Dolomite and Kupferschiefer. Whereas X-Ray Transmission and X-Ray Fluorescent sensors are able to distinguish between low grade and higher grade ores. None of them is capable of sorting by lithology. This chapter provides an overview of the physical and technical application principles of NIR Sorting in the mining industry. It includes the preceding investigations that demand for medium scale sorting tests. The last part of this chapter presents all results of the test and its conclusions.
PRINCIPLES OF NEAR INFRARED SORTING
NIR sorting is based on NIR spectroscopy. With spectroscopic techniques the interaction of electromagnetic radiation with a molecular system are studied. It is possible to derive physical and chemical material properties on a very detailed level (Clarc, 1999). NIR spectroscopy has been used as a rapid and inexpensive technique for material analysis in the laboratory for over seventy years and is one of the most important analytical techniques
available to today´s scientists (Robben, 2010). In the case of mineral NIR Sorting the reflectance of a NIR radiation is sensed and evaluated. At certain wavelengths, the radiation in the NIR spectrum excites molecules to oscillate which is therefore absorbed, while another part of the radiation is reflected and redirected to a detector unit. The reflected intensity plotted over the wavelength results in an individual spectrum, providing information about the chemical composition and purity.
A major amount of minerals can be classified based on their spectral response in the NIR and most diagnostic absorption features occur in the region between 1300 nm and 2550 nm (Clarc, 1999). The source of radiation can either be placed above the conveyor belt or above a declining chute scanning the material in near free fall. A simplified assembly of a sensor based sorter and all of its components is displayed in Figure 5.
The source emits the NIR signal, which is partly absorbed and partly reflected by the material on the conveyor belt. The reflection is detected and evaluated instantly and automatically by the Data Processing Unit. Before the NIR Sorter can function automatically an extensive calibration for the specific sorting task has to take place. During the calibration several spectra of the different groups that are supposed to be separated have to be
recorded (refer Figure 4). Groups can be, for example different materials (plastic & rubber) or, as described in this article, different rock types/ lithologies. In a second, step the operator uses the sorting software to form the specific spectrums´ characteristics into classified groups. During the actual sorting application the data processing unit assigns each scanned reflection to one of the classified groups. The operator may decide which group shall be rejected and which shall be accepted. The reject fraction is hit by a blast of compressed air and sent into a different process stream. The accept fraction remains untouched. In general the reject fraction is smaller in terms of mass than the accept fraction. NIR Sorters can be implemented in mineral processing to fulfill various tasks. They can be applied in rougher or preconcentration stages for coarse size ranges (up to 250 mm) down to concentration stages for particle sizes down to 1 mm. Below this size, sensor based sorters in general are usually not economic because of the correlation between average feed size and throughput (Wotruba, 2011). For the application of NIR sorting of Kupferschiefer ore a particle size of 20 to 50 mm has been determined.
CONCLUSION
The recording of the NIR spectra show significant differences in the shape of the spectrum as well as the intensity of the the reflection. The NIR Sorting test results proof, that a separation of the sandstone lithology is possible. With a clean rock surface and under wet conditions the sorter is able to derive sufficient sorting characteristics from the detected NIR spectra to distinguish between the sandstone and the dolomite-Kupferschiefer ores. The sorter rejects the sandstone, which has a mass yield of 20 %. The recovery of the sandstone reaches up to 95 % with less than 2 % of the total organic carbon.For further mineral processing this sorting output would represent a separate process stream for flotation. The evaluation of the test results show that a majority of the sorting mishits that cause sandstone recovery loss and impurities of dolomite and Kupferschiefer in the sandstone fraction are due to inaccurate pressure air blasts and detection errors through overlapping. Both these causes could possibly be improved in a sorting application.
Further flotation work on the sandstone and dolomite-Kupferschiefer fraction is necessary to evaluate the impact of the NIR Sorting on the complete mineral processing line of Kupferschiefer ore.
Add to My Watchlist
What is My Watchlist?
