Industrial mineral sorting

Quality control of rock
Quality control of rock, stone & minerals

Minerals, stone and rock

The production of high-quality mineral products and building materials demands high-grade raw materials as well. Expensive and time-consuming laboratory methods are usually used for the analysis of raw materials. Inline analysis is commonly realised by RGB cameras. For a successful separation by RGB cameras, the colour of the valuable minerals and the waste rock fraction must differ significantly. If both fractions show a similar colour, the application of NIR-process technology will enable a successful separation nevertheless. Each NIR active mineral shows a unique pattern in the NIR spectrum. These spectral features are used as basis for the identification routines in mineral sorting. The identification routines are developed specifically for each sorting task, depending on the mineral composition. The routines can be adjusted easily by the customer.

The imaging NIR process spectrometers of the uniSPECx.xHSI series are ideally suited for such identification tasks. Due to the high spatial resolution, even small particles are detected at a conveyor belt speed of up to
3 m/s. The multiplexed NIR-spectrometer uniSPEC2.2MPL-24V is used for analysis of large particles. All NIR-systems are easy to integrate into existing plants of the process industry or in the quarry. For the separation of valuable minerals from waste rock, several identification routines are available. An overview for standard applications is given in the table below.

Application uniSPEC2.2MPL-24V uniSPEC1.7HSI uniSPEC1.9HSI uniSPEC2.2HSI
Calcite / Dolomite + - - ++
Talc / Waste rock + + + ++
Borate / Waste rock + ++ - -
Bauxite / Waste rock + - - +
Gypsum / Anhydrite + + ++ -

++ particularly suitable  + suitable  - limited suitability

The NIR-process spectrometers uniSPECx.xHSI and uniSPEC2.2MPL-24V can be used for semi-quantitative determination of minerals in rock. The results can be used to comply with product specifications and limits.

Qualitative separation of calcite and dolomite

Sorting of limestone by NIR technology
Sorting of limestone by NIR technology

The NIR cameras uniSPEC2.2HSI and uniSPEC2.2HSIsens in combination with module idMgO

Limestone has a global annual production rate of 1000 mega tonnes and is therefore one of the main products in the mining industry next to sand and grovel. Limestone must contain a preferably low amount of magnesium oxide (MgO) for usage in cement production or as additive in smelting. The amount of MgO often originates from dolomite impurities in the raw material.

The NIR application idMgO was developed for the identification of MgO phases in limestone. This application significantly reduces the amount of MgO in the resulting product and is customised for specific on-site conditions of the natural mineral deposit. A RGB-camera can be combined with the NIR cameras uniSPEC2.2HSI and uniSPEC2.2HSIsens for additional identification of differently coloured impurities.

Qualitative separation of spodumene

Identification of waste rock in spodumene
Identification of waste rock in spodumene

The NIR camera uniSPEC1.7HSI or multiplexed NIR spectrometer uniSPEC1.9MPL-24V in combination with module idSpodumene

Spodumene is the most commonly exploited lithium mineral. The industrially important lithium compounds are gained by energy-consuming mineral processing. Especially the initial processing step requires temperatures of up to 1000 °C. An accumulation of the spodumene by reduction of the waste rock content in the initial feed is therefore crucial for the optimisation of energy and chemicals consumption during the process.

Common minerals accompanying spodumene in deposits are feldspars and quartz. These minerals exhibit similar colours to the white or yellow spodumene, which hinders a successful separation of the waste rock by colour sorting. The optical sorting by NIR can be applied nevertheless. Due to the mineralogical differences between spodumene and quartz / feldspar, an identification of spodumene is possible.

The identification module idSpodumene was developed for identification of waste rock containing quartz or feldspar in spodumene deposits. The module is available for uniSPEC1.7HSI hyperspectral imaging cameras and suitable for sorting applications (grain size fractions ≥ 40 mm).

Input material: Mixed spodumene rock
Valuable fraction: Spodumene
Waste rock: Feldspar, Quartz
Particle size: > 200 mm (MPL-24V), > 40 mm (HSI)
Conveyor belt width: 800 mm up to 1600 mm
Conveyor belt speed: Up to 3 m/s

Videolink "Identification of waste rock in spodumene"

Qualitative separation of quartz, calcite and clay

Identification of waste rock in quartz
Identification of waste rock in quartz

The NIR camera uniSPEC2.2HSI or multiplexed NIR spectrometer uniSPEC2.2MPL-24V in combination with module idQuartz

High purity quartz is crucial for several applications such as glass production or preparation of silicon for electronic components. Differently coloured impurities can be identified successfully by common RGB cameras but similarly coloured impurities such as limestone or clay minerals still remain in the product.

NIR spectrometers such as uniSPEC2.2HSI can identify similarly coloured limestone or clay minerals in quartz due to the chemical and mineralogical differences. The application idQuartz was developed for the NIR hyperspectral camera uniSPEC2.2HSI, enabling the identification of calcite and clay mineral impurities in quartz for sorting applications (grain size fractions ≥ 40 mm).

Input material: Mixed quartz rock
Valuable fraction: Quartz
Waste rock: Calcite, clay
Particle size: > 200 mm (MPL-24V), > 40 mm (HSI)
Conveyor belt width: 800 mm up to 1600 mm
Conveyor belt speed: Up to 3 m/s

Quantitative determination of MgO in limestone

Quantitative determination of MgO in limestone
Quantitative determination of MgO in limestone

The MgO content in limestone must be determined regularly during cement production. For this purpose, XRF measurements are performed routinely on powdered samples, although the sample preparation is costly and time consuming. In contrast, quantitative NIR-measurements of the MgO content using NIR spectrometer are possible without further sample preparation.

The quantitative analysis of the MgO content is based on the identification routine idMgOquant. This routine is available for the hand scanner of the uniSPEC2.2MPL-24V process spectrometer for at-line analysation. Measurements on reference samples with known MgO content show a high correlation of NIR-analysis results and XRF data.

Analysis of minerals, stone and rock by NIR hyperspectral camera