Metso Insights Blog Mining and metals blog Ore sorting the road to optimizing your operation
Mining Metals refining
Oct 16, 2017

Ore sorting: the road to optimizing your operation

Sensor-based ore particle sorting offers a powerful method for pre-concentration or waste rejection early in the comminution process. The technology analyzes each and every rock particle on-line for unique physical and chemical properties and separates the particles by high pressure air jets. This maximizes recovery or upgrade and is suitable for various minerals and applications. Successful implementation relies on ore heterogeneity, differential sensor response and suitable feed preparation.
Outotec-Tomra sensor-based ore particle sorters
Figure 1. Outotec-Tomra sensor-based ore particle sorters. Machines with X-Ray transmission sensors (on left) and optical/laser sensors (on right) are available, as well as electromagnetic and other specializes sensors.

Sensor-based sorting technology

Sensor-based sorting has been around since the 1970’s. In the early days mainly radiometric and photometric sensors were used to distinguish characteristics of materials. As it is common with new technology it was very expensive, and computing performance was highly limited making it often not economic viable. The information revolution at the turn of the millennium boosted computing power, making it cheap and available. Today data reconciliation, validation and analyzing does not limit the sorting process any more. The focus on research and development is now on new sensors and sensor accuracy with the aim of improving detection accuracy.

Outotec provides pneumatic sensor-based sorting machines for pre-concentration in the mining and minerals processing industry, via our partnership agreement with Tomra Mining. Different sensor and machine types cover a vast selection of ore and mineral types. All machines are built with the same three principles: material presentation, data collection by sensors, and separation. Material is presented to the sensors as a monolayer of liberated particles. Different sensor types analyze every single particle, and the decision of accepting or rejecting is made within 30 milliseconds. Physical separation of particles is achieved by altering the particle trajectory with precise air jets into the product or waste chutes.

The principles of sensor-based ore particle sorting
Figure 2. The principles of sensor-based ore particle sorting.

Benefits of sorting

Sorting enables the removal of sub-economic waste material in a very early process stage. When sorting is situated in the crushing circuit, the removed waste does not have to be milled further down and processed in the beneficiating plant. The removed material allows for more throughput of valuables in the downstream process, or major capital investment savings by decreasing plant size. This decreases the use of consumables such as water or electricity per unit of metal produced. A sorting machine can also be placed in the pebble circuit of an AG or SAG mill. Removing waste from the milling circuit can de-bottleneck many processes.

With the addition of sorting to the processing plant, significant economic opportunity exists in optimizing mining with reference to pre-concentration. The resource size can be increased by considering a lower cutoff grade and less selective mining techniques that offer a lower operating cost per tonne. The additional waste material mined due to less selective mining is removed by sorting and the feed grade to the plant remains high. In addition, flotation and leaching recoveries are generally increased due to the higher head grade.

Many mining operations accumulate stockpiles of material below the economic cut-off grade threshold. Numerous successful test work and installations have been done on reclaiming waste dumps by upgrading the low grade material. Applying sorting to such material has low operating costs because the material has already been blasted and mined.

The benefits of sensor-based ore particle sorting
Figure 3. The benefits of sensor-based ore particle sorting.


Every material is unique. The sortability of any given deposit is dependent on the heterogeneity of the ore. Any successful sorting operation requires differences in some material properties. These properties may be color, density or electro-magnetic. Different mineralizations can also be distinguished based on near infrared spectra or differences in the crystalline structure. In some cases the disseminated hints of metal are not visible to any sensor and sorting relies on tracer properties or mineral associations. This is especially true for gold deposits, where gold particles within the rock are almost always below the resolution of the machine (<1 mm pixels). To verify the correlation of tracer properties and the metal grades test work has to be done for every sorting application prior to implementation.

Sorting test work

Sorting test work starts on the desktop with a sorting amenability evaluation. Experts validate all data available and provided by the customer. Based on expertise and prior test results a first gate decision is made to continue or not.

Next is the first stage of hands on sensor test work on the material. Outotec asks for 10-20 carefully selected samples of each rock type representing the whole ore body for a first inspection test. These rock types may consist of: ore, marginal ore, waste, intermediate waste or SAG pebbles. Testing can be completed at the Tomra Mining test centers in Hamburg, Germany and Sydney, Australia. Responses from different sensors to the material is tested to determine differences between rock types. The aim of this test work stage is to provide sufficient information for initial sensor selection.

After a successful first inspection a performance test is carried out on industrial scale machines. Test sample sizes is typical 1-4 tons of material for each sample. The sample should be as representative as possible of the expected feed including a mix of all rock types.

After assaying the results this test provides all necessary information for proof of sortability and expected performance. During the performance test the sensor system is adjusted and calibrated according to the requirements of the ore body. This provides crucial information such as grades, recovery and capacities. More complicated flow sheets including scavenger sorters or cleaner sorters or combining different sensor types can be tested as well.

Flowsheet of test work for sensor-based ore particle sorting
Figure 4. Flow sheet of test work for sensor-based ore particle sorting.

Process design

A successful performance test campaign provides all relevant information for process design and implementation of a sorting solution. Test work is completed using industrial sized machines, so therefore no scale-up factor has to be applied and piloting is generally not required. The main focus of process design is the pretreatment of the sorter feed. The key to successful operation is sufficient fines removal and a narrow particle size distribution. To maximize the benefit of sorting, the circuit should be designed to deliver the maximum amount of material in the size range for sorting. The optimal location for a sorting solution is within the secondary crushing stage where a minimum amount of material bypasses sorting through under- or over-size streams.

Capacities for individual machines varies based on the rock type and can be up to 400 tons per hour for certain applications. By combining various machines in different particle size ranges plants have been implemented with a maximum capacity of 2000 tons per hour total plant feed. Outotec’s expertise in crushing, grinding, materials handling, flotation, leaching and dewatering ensures smooth incorporation of sorting in any given process. Our offering includes everything from sorting equipment packages including auxiliaries to full turnkey solutions.

Optimized Operations

Sensor-based ore particle sorting can be used to pre-concentrate ROM ore feed, SAG Mill scats and low grade or waste stockpiles. Plant throughput increase, reduced operating costs and mine optimization should be considered in evaluating sorting as well as increased recovery in flotation and leach circuits. The economic benefit to the operation can be very powerful and also typically reduces sensitivity to metal prices. A recently published example is Minsur’s San Rafael mine in Peru, showing a pay-back period of only 4 months after the start-up of the stockpile sorting plant. "The machines have proven to be very efficient," said Pedro Condori, Process Operation Manager for Minsur.

{{zoomedImageIndex + 1}}/2
{{ captions[zoomedImageIndex] }}
Mining Metals refining