Available via license: CC BY 4.0
Content may be subject to copyright.
© The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution
License 4.0 (http://creativecommons.org/licenses/by/4.0/).
E3S Web of Conferences 18 , 01001 ( 2017 ) DOI: 10.1051/e3sconf/20171801001
MEC2017
* Corresponding author: jacek.kolacz@comex-group.com
Advanced separation technologies for pre-concentration of
metal ores and the additional process control
Jacek Kolacz1,*
1Comex AS, Box 53, 1309 Rud, Norway
Abstract. A new sensor based sorting system has been developed at Comex, based on complex image
analysis. The sorting system employs different images generated by visible light and X-ray to provide
maximal information about processed particles. Application of X-ray makes it possible to analyse the
internal structure of the particles, which provides a 2D image with the valuable information. The
advanced system for the texture and the pattern recognition is applied to analyse these images
simultaneously, to provide the required identification of the particles for efficient separation. The
paper describes a number of tests during sorting of Cu-Zn-Sn ores with high separation efficiency. As
an example, the Cu content is enriched from 0.4% to 1.25% and the Zn content is increased from
0.83% to 2.24%. In addition, the performance of the X-ray sorting system is described for enrichment
of Zn-Pb ore and compared with the heavy media separation technology. Simplifying the pre-
concentration through the use of the sorting technology can significantly reduce the cost of these
operations. Finally, the operating data from the sorting system have been analysed regarding
a potential use for controlling the plant operation. This can bring the new tools for control strategies
for the complex processing plants.
1 Current technology
Processing of metal ores including zinc and lead, is
mainly based on gravity separation equipment. It often
includes applying dense media separation systems,
especially for large particles. Such equipment provides an
efficient separation, however, it has a number of
disadvantages and limitations. The most important of
them can are listed below:
• It is necessary to apply complicated media
conditioning equipment to achieve a stable liquid density,
like pumps, dosage systems, mixers, density sensors, etc.
• It requires applying heavy density materials like
ferrosilicon, which are expensive.
• The high density materials must be washed thoroughly
from the particles to avoid loses and limit product
contamination.
• The system requires use of water, which must be
purified and recycled. It also requires heating in low
temperature environments.
• The washing system is mechanically complicated with
many moving parts, which require servicing and
maintenance.
• The complete system requires a lot of space in form of
both height and foot print area.
All these disadvantages provide a high cost of
operation [1] especially for the low production capacity.
It is very often not feasible to apply such systems for low
or medium production scale. Consequently, it is necessary
to build the central processing plants, which require more
transportation lines and complicated logistics.
Finally, the dense media separation process provides
a relatively low separation sharpness. The particles, which
have very high or very low density, when compared to the
suspension density, will very quickly sink or float, thus
providing a good separation effect. However, when the
separated material has a density closer to the suspension,
the sinking or floating effect will require more time and
this can create more chances for misplaced particles
during this process. Such effect is not taking place when
sensor based sorting is applied.
2 X-ray sorting technology
Application of the sensor based sorting systems, without
any doubt, brings a lot of possible advantages in many
mineral operations. The main advantage of the new
system is related to its universality and at the same time
very sophisticated image processing functions, which can
be carried out in the same processing unit [2]. Figure 1
shows the CXR-1000 system from Comex with 1 m belt
width, in the configuration where many different analysed
parameters can be used to provide particle separation. The
image analysis system includes a camera installed either
over the transport belt conveyor or at its discharge end.
The system includes the X-ray attenuation analysis
realized by the XRT system in the central part of the
conveyor belt. The sorting system can be used with both
2
E3S Web of Conferences 18 , 01001 ( 2017 ) DOI: 10.1051/e3sconf/20171801001
MEC2017
optical and XRT analysis or separately depending on an
application.
Particle recognition used to separate different
materials is based on a complex shape and colour analysis
where the particles can also be identified by over 20
parameters used for shape description. Some of them are:
diameter in different orientations, perimeter, centre of
mass, moment of inertia, particle elongation factor, edge
sharpness, etc. Additional combinations of these
parameters can also be used for distinguishing particles of
interest. The surface of particles where different colours
or contours vary in intensity and frequency can be
analysed by FFT filtration (Fast Fourier Transformation)
to recognize differences in texture and structure of the
processed particles. This analysis brings much more
complex information about the analysed particles rather
than colour recognition alone. Finally, the XRT picture is
integrated into the optical analysis, which provides a lot
more information about the particle surface properties and
its internal structure. All these sophisticated analysing
functions require a lot of computation power and they
have to be optimized to allow high capacity sorting. This
is done by special program architecture and algorithm
solutions allowing efficient management of the
calculation routines and sorting priorities. This allows
achieving still high separation capacity and extremely
high efficiency [2], where the product purity can reach
even 99.9 %. Such results can be achieved in different
production scale depending on the process requirements
[3].
The sorting system as shown above can be configured
with different sensors depending on the particular
application. In the test work as described in this paper,
only the X-ray sensor was applied since the optical
properties did not show any possibility for differentiation
of particles of interests. The X-ray sensor provided
possibility to calculate a density of separated particles.
Therefore, it was possible to compare this new sorting
technology with dense media separation as the traditional
processing method in this process. Such sensor based
sorting systems are already successfully tested and
applied in the metal ore processing like iron ore [2]
including Zn/Pb ore [4] and coal [5].
Fig. 1. Comex CXR sorting system – configuration with the X-
ray and the optical analysis.
Application of the sensor based sorting is very
attractive when compared to other separation
technologies. It provides the following advantages, when
compared to the DMS systems:
• Possibility to apply in small capacity production lines
• Small space requirement
• No use of water (it is completely dry technology)
• No complicated additional equipment like, pumps,
mixers, conditioners, recuperation devices,
instrumentation, etc.
• No requirement of high density powder (ferrosilicon)
• Lower investment cost (about 25-50%)
• Very low operating cost (about 15-20 time lower)
It was therefore very interesting to compare the
separation efficiency using the similar input material.
3 Comparison of the DMS and X-ray
separation results
The DMS and the X-ray sorting system have been
compared on the base of the same feed material. The test
work was performed in the zinc and lead mine DPM
„Olkusz – Pomorzany” in Poland, where the DMS system
was applied together with flotation as the main
enrichment process [4]. The DMS system was employing
the Wemco unit operating in 20-60 mm size range. It was
possible to sample the input and outlet fractions and carry
out the necessary chemical analysis and gravity separation
tests. The samples included few hundred kg material
fractions, taken under the stable operating conditions of
the DMS system. The results of the detailed analysis of
the feed, sink and float fractions are shown in Table 1.
Table 1. Separation results in the Wemco DMS system.
Fraction
Rec.
[%]
Grade [%]
Recovery [%]
Zn
Pb
Fe
Zn
Pb
Fe
Feed
100
1.80
0.80
4.17
100
100
100
Sink (heavy)
19.13
7.63
3.93
16.80
81.31
93.77
76.77
Float (light)
80.87
0.41
0.06
1.20
18.69
6.23
23.23
Fig. 2. Picture of the DMS outlet fractions together with the X-
ray image of the same particles.
In addition, the sink and float fractions, were analysed
by the Comex CXR sorting unit to investigate the content
of high and low density particles. Figure 2 shows these
fractions in form of the optical picture and the X-ray
image. The blue colure shows the low density particles
and the red and yellow colours indicate the high density
material. It can be noticed that the sink fraction still
contains the low density particles which can be recovered
by the sorting unit and separated to the waste fraction. It
was therefore concluded that the separation process in the
3
E3S Web of Conferences 18 , 01001 ( 2017 ) DOI: 10.1051/e3sconf/20171801001
MEC2017
X-ray sorting unit might provide better result in form of
increased recovery or grade values.
The Comex CXR-1000 X-ray sorting unit with 1 m
belt width was used for testing. It was fed with the same
material as the feed sample taken from the DMS unit. It
was therefore possible to compare the separation effect
from both separating units using the identical feed
material composition. The separation results are shown in
Table 2.
Table 2. Separation results in the CXR sorting system.
Frac.
Rec.
[%]
Grade [%]
Recovery [%]
Zn
Pb
Fe
Zn
Pb
Fe
Feed
100
1.80
0.80
4.17
100
100
100
High dens.
10.96
5.42
15.36
79.16
87.12
53.40
10.96
Low dens.
87.00
0.43
0.12
2.00
20.84
12.88
46.60
When comparing the results in Table 1 and 2, it can
already be noticed that the a different enrichment level of
investigated minerals was registered for various
recoveries. For better illustration, the rade and the
recovery points were shown on Figure 3.
Fig. 3. Grade-recovery illustration for the DMS and X-ray
sorting processes.
For the DMS process, the Pb content was with lower grade
but higher recovery, when compared to the X-ray sorting.
However, this can be adjusted by setting of the sorter to
achieve a similar level of both parameters. Regarding the
Zn separation, there is a significant difference in grade for
the similar recovery level, where the X-ray sorting
provides almost 11% grade, while DMS only 7.6%, for
the similar recovery of about 80%. Finally, the recovery
of Fe is much lower for and the grade is reduced, which is
beneficial for this particular process of Zn-Pb production.
4 Other sorting applications
Another application of the CXR sorting unit is shown for
the Cu-Zn-Sn ore in one of the mines in China. In this case
the ore was investigated in two main material streams
where the Cu concentration was the main component in
the first case, and the second stream contained mainly Zn
and Sn. In this case the material was processed by the
external processing plant where the large quantities of the
ore had to be transported with significant distances. The
sorting solution with the CXR system was suggested to
provide improvements in the overall process. The
separation results are shown in Table 3 for the Cu ore. In
this case, the feed material containing 0.404% of Cu was
enriched to 1.258% with 85.65% recovery of Cu. At the
same time the concentrate stream after sorting system,
corresponded to only 27.52% of the incoming mass. It
means the pre-concentrated material was significantly
reduced in quantity before further processing.
Table 3. Cu ore separated in the CXR sorting system.
Fraction
Yield
[%]
Density
[kg/l]
Test Elements [%]
Cu
Zn
Sn
Cu Conc.
27.52
3.09
1.258
0.989
0.299
Cu Tailing
72.48
1.92
0.080
0.151
0.096
Cu Feed
100
2.24
0.404
0.382
0.152
Fraction
Yield
[%]
Density
[kg/l]
Recovery [%]
Cu
Zn
Sn
Cu Conc.
27.52
3.09
85.65
71.32
54.18
Cu Tailing
72.48
1.92
14.35
28.68
45.82
Cu Feed
100
2.24
100
100
100
Table 4. Zn/Sn ore separation in the CXR sorting system.
Description
Yield
[%]
Density
[kg/l]
Test Elements [%]
Cu
Zn
Sn
Zn/Sn Conc.
28.30
2.87
0.145
2.249
0.655
Zn/Sn
Tailing
71.70
2.65
0.023
0.273
0.097
Zn/Sn Feed
100
2.71
0.058
0.832
0.255
Description
Yield
[%]
Density
[kg/l]
Recovery [%]
Cu
Zn
Sn
Zn/Sn Conc.
28.30
2.87
71.33
76.48
72.72
Zn/Sn
Tailing
71.70
2.65
28.67
23.52
27.28
Zn/Sn Feed
100
2.71
100
100
100
The similar results were achieved for the Zn-Sn ore as
shown in Table 4. In this case, the feed contained 0.832%
of Zn and 0.255% of Sn. After sorting, the metal content
was increased to 2.249% of Zn and 0.655% of Sn with the
recoveries of 76.48% and 72.72% correspondingly. At the
same time, the material quantity was reduced to 28.3% of
the total incoming mass to the sorting process. Again, this
provided huge savings in terms of the further processing
cost.
5 Operating cost improvements
Sorting systems have the important limitation related to
the capacity of the single processing unit. This is related
to the operating principle where the sorting process is
carried out in the mono-layer of the particles introduced
to the sorting unit. It means that higher throughputs must
0
10
20
30
40
50
60
70
80
90
100
01020
Recovery [%]
Grade [%]
Zn DMS
Pb DMS
Fe DMS
Zn CXR
Pb CXR
Fe CXR
4
E3S Web of Conferences 18 , 01001 ( 2017 ) DOI: 10.1051/e3sconf/20171801001
MEC2017
be provided by the parallel configuration of the sorting
units. Dry X-ray separation equipment provided by
Comex fulfils this main cost requirement for multiple
stage separation keeping the total cost at a competitive
level. The resent development and progress in
microelectronics provided quite a number of new possible
solutions within automation and separation techniques.
Comparison with the traditional processing methods like
DMS, indicate very clearly that the dry X-ray sorting
systems can be very competitive regarding investment
cost, operating cost and other environmental aspects like
no need of water. Simple comparison of the DMS
washing plant and the sorting plant based on CXR Comex
sorters, is shown in Table 5.
Table 5. Cost comparison between the traditional washing
plant and the Comex dry beneficiation plant.
Option
DMS
plant
Comex dry
sorting plant
Processing
capacity
[MTPA]
4.5-5
4.5-5
Water
consumption
[l/t]
1-2
0.01
Investment cost
[USD]
15 mln
10-12 mln
Operating cost
[USD/t]
1-1.5
0.05-0.1
The traditional DMS plant is characterised by a simple
operating principle, however, it requires a lot of moving
parts, motors, belts, pumps, magnetic recuperation as well
as access to water and heavy media like ground magnetite
or ferrosilicon. This result in high both investment and
especially operating cost. Based on the average plant size
providing 5 MTPA the investment cost is about 15 mln
USD and the operating cost ranges about 1-1.5 USD per
ton. In addition there is a cost related to water supply
which is very difficult to calculate especially for hot
climate countries.
For the alternative process where the CXR sorters are
employed, the separation can be done in combination of
parallel-serial connections of the sorting units and the
necessary screening equipment. The investment cost
ranges between 10-12 mln USD and the operating cost is
about 0.05-0.1 USD per ton. It means the operating cost
of the X-ray sorting plant is about 15-20 times lower then
the corresponding process employing DMS. The benefit
of using the dry X-ray sorting plant is growing rapidly
when the processing capacity is lower. For smaller mines,
it is often not feasible to build the washing plant while the
dry sorting system having smaller capacity, can easily be
applied.
6 Data processing potential
The sorting system is usually installed at the high
throughput streams, which are at the early stage of
processing. It means many particles are passing the
sensors of the sorting equipment and these particles very
often correspond to the ROM material or the main
material introduced for the enrichment process. By
collecting this data on-line, it is possible to control the rest
of the process by adjusting its parameters to obtain the
optimal conditions and adjusting the flowsheets streams.
The control system can adjust e.g. the reagents
concentration used in floatation based on the information
about the metal or rejects concentration in the ore. It is
also possible to use a fuzzy logic control to adjust the
processing paths in the flowsheet to get the best
performance of the plant. It brings new possibilities for
optimising the complex processing plants. This
knowledge has never been used before and it will be
further investigated in the future work.
7 Conclusions
The CXR X-ray sorting unit can provide a similar or even
better separation effect than DMS system, during pre-
concentration of the Zn-Pb ore before further processing.
The similar separation process can be realised by X-ray
sorting as a much simpler system, regarding lower
number of moving parts, lower complication level, as well
as reduced maintenance and servicing. No need of water
and the high density weight material, can be important for
more flexible and more environmentally friendly
operation. Finally, in terms of the reduced investment and
operating cost (by 15-20 times), it can dramatically
improve the mine operation regarding economic aspects.
Finally, the sorting results can be transferred to the
valuable data describing the input material. Such
information can be used for controlling the rest of the
processing plant by adjusting the control models for
flotation regents and different flowsheet configurations.
References
1. Wills B.A., Wills' Mineral Processing Technology:
An Introduction to the Practical Aspects of Ore
Treatment and Mineral Recovery (2006)
2. Kolacz J. Sensor based sorting with signal pattern
recognition: the new powerful tool in mineral
processing (XXVII International Mineral Processing
Congress, Santiago-Chile,106-115, 2014)
3. Kolacz J. Efficient scale up of sensor based sorting
systems. XXVIII International Mineral Processing
Congress, Quebec, Canada, 2016)
4. Wieniewski at al., A. Wieniewski , E. Szczerba, A.
Nad, R. Luczak, J. Kolacz, Szewczuk, Evaluation of
the application possibilities of modern separation
techniques for pre-concentration of the Zn-Pb ore,
(XI International Non-Ferrous Metals Conference
ICNOP, Trzebieszowice, Poland, 2015)
5. Kolacz J. Dry coal beneficiation by sensor based
sorting (Mineral Engineering Conference,
Szczawnica, Poland, 2015)