Professional Documents
Culture Documents
Beneficiation Routes For Upgrading Iron Ore Tailings With A Teetered Bed Separator
Beneficiation Routes For Upgrading Iron Ore Tailings With A Teetered Bed Separator
To cite this article: O. Ozcan & I.B. Celik (2016) Beneficiation routes for upgrading iron ore tailings
with a teetered bed separator, Separation Science and Technology, 51:17, 2844-2855, DOI:
10.1080/01496395.2016.1218514
Beneficiation routes for upgrading iron ore tailings with a teetered bed
separator
O. Ozcan and I.B. Celik
Department of Mining Engineering, Hacettepe University, Ankara, Turkey
CONTACT O. Ozcan ozgurozcan@hacettepe.edu.tr Department of Mining Engineering, Hacettepe University, 06800 Beytepe, Ankara, Turkey.
Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/lsst.
© 2016 Taylor & Francis
SEPARATION SCIENCE AND TECHNOLOGY 2845
60
to dry and wet process options are proposed.
50
40
Materials and methods 30
20
The representative samples were obtained from the
stockpiled tail of the dry low intensity magnetic 10
Sample preparation
It is known that TBS separates the particles based on
both particle size and density. In addition, if particles
are fed to a hindered settling classifier in narrow sizes,
the effect of size over density is eliminated, and besides
classifying, beneficiation will also be enhanced.[19] With
this fact and considering the liberation data narrow
90
80 -1.00+0.50 mm
70 -0.50+0.10 mm
60
Amount (%)
50
40
30 Figure 5. Schematic view of sample preparation.
20
10 Table 1. Weight and iron distribution of the head sample.
0 Sample Weight (%) Fe (%) Recovery (%)
95<x 75<x<95 50<x<75 0<x<50
(Liberated) −1.00 + 0.50 mm 50.50 28.80 60.45
Liberation Class −0.50 + 0.10 mm 42.43 19.65 34.65
−0.10 mm 7.07 16.68 4.90
Figure 4. Liberation degree of hematite. Head sample 100.00 24.06 100.00
SEPARATION SCIENCE AND TECHNOLOGY 2847
tory scale cross flow design with dimensions of 100 mm mono particle layer. While the splitter angle was set to
× 100 mm cross section and 500 mm high (square tank obtain a high grade concentrate, roller speed variation
height), followed by a 200 mm conical section. (126, 201 and 276 rpm) was examined whether it has an
The main aim of the rougher beneficiation study was effect on grade and recovery.
to produce a clean tail (TBS overflow) and a high iron Slime fraction was also subjected to the wet high
recovery for concentrate (TBS underflow). To obtain an intensity magnetic separator to check if any upgrade
overflow product, the teeter water rate (12 lpm) was kept is possible. These series of tests were conducted by
at maximum for coarse fraction. The teeter water rate varying the magnetic intensity, whereas the other vari-
changed as 4 and 7 lpm for fine fraction. In addition, ables such as the feed rate of dry solids (0.6 kg/hour),
the pressure set point was also set to a relatively high value pulp density of solids by weight (20%), and wash water
to create the beneficiation effect because, it is known from flow rate (5 lpm) were kept constant.
the literature, at lower pressure set points, the TBS merely
acts as a size classifier[20] and the beneficiation effect of
the teetered bed disappeared. Samples from the feed,
Results and discussions
overflow and underflow streams were collected at steady
state conditions. Gravity concentration was performed on In all beneficiation options, the upstream operation
the samples by a laboratory scale shaking table (500 × involved classification by screens. Only for the wet
1200 mm). To obtain a high grade concentrate, wash process was a rougher stage by TBS performed after
water rate (15 lpm) and tilt angle (5°) were kept at max- screening to see its impact on beneficiation. The under-
imum during the test. Two different concentrates, mid- flow of the TBS was concentrated in two routes utiliz-
dling and tail samples, were collected for each test. ing the difference of valuables and the gangue in terms
As an alternative to gravity concentration, magnetic of gravity and magnetic susceptibility. Besides gravity
separation was also conducted to the underflow pro- and magnetic separation as wet methods, a dry option
duct of the TBS by using a laboratory scale WHIMS. In of the flow sheet was also considered.
the present work, magnetic field intensities were set as
6960, 8625 and 9630 gauss, respectively. The effect of
magnetic field intensity, and magnetic susceptibility
Dry beneficiation by a high intensity magnetic
were studied simultaneously.
separator (Permroll separator)
Magnetic separation of samples was also examined on
a dry basis. To propose an alternative flow sheet on dry In this study, the effect of roller speed and particle size
mode both fractions were upgraded by a high intensity on the iron grade and recovery was investigated.
permanent magnetic roll separator (PERMROLL), which Sufficient magnet groups were used for coarse and
has a 6000 gauss magnetic field strength. In the experi- fine size fraction separately. The test conditions and
ments, the feeding rate was kept constant to create a results are given in Table 2.
2848 O. OZCAN AND I. B. CELIK
The results indicate that the coarser size fraction Rougher beneficiation by TBS
(−1.00 + 0.50 mm) can be separated by this technique.
The geometrical properties of the TBS allow the creation
It is possible to increase the iron content of the tail
of a highly dewatered underflow product. The underflow
from 28.80% to 66.40% Fe for coarse fraction. However,
solid content of coarse fraction was determined to be
fine fraction can only be upgraded from 19.65% to
75.68% while the underflow solid contents of fine fraction
42.08% Fe.
were determined to be 74.25% and 75.48% for a low teeter
The roller speed in the Permroll determines the
water rate and a high teeter water rate, respectively. The
centrifugal force which in turn determines the grade
chemical analyses of the TBS products are given in Table 3
and mass yield to the concentrate, which means
for coarse and fine size fractions.
increasing the roller speed leads to only high mag-
As seen from Table 3, the underflow stream was
netic particles to concentrate. Improvement in the
concentrated up to 56.95% Fe from the feed value of
quality of the concentrate was achieved for both
28.80% Fe for coarse fraction. The Fe content of the
size fractions by increasing the roller speed, but the
overflow in the coarse fraction was only 3.44% and
target grade could be achieved at a coarse size frac-
Fe recovery in the underflow was recorded to be a
tion only. This could be attributed to the feed iron
considerably higher value such as 93.72%. The results
contents of these two different size fractions. Heavier
indicate that a significant upgrade could be achieved
coarse particles were affected more by the centrifugal
with the coarse size fraction in which the specific
force created by the roller and non-magnetic coarse
gravity differences of iron minerals and gangues are
particles can be easily reported to tail. However,
more pronounced. The relative settling velocity of
some weak magnetic and locked particles also report
particles depends on the size, shape and specific
to tail because of their weight and increased tail iron
gravity. If any two factors of the above are uniform,
content. Tail grade of coarse fraction at 276 rpm was
the settling velocity of particles depends only on the
the maximum. This is a clear effect of a high cen-
specific gravity. For example, if the particle has a
trifugal force. Fine fraction could not be influenced
narrow size distribution and the same shape in a
as much as the coarse fraction and some of the
given liquid the relative settling velocity is dependent
weakly magnetic particles reported to concentrate
on the specific gravity differences of particles. So, the
and reduced the iron content. The recovery of the
upward water can only report less dense gangue
concentrates decreased with the increase in roller
minerals to the overflow stream.[16]
speed for both size fractions as expected.
The results of fine fraction can be compared in terms of Table 4. The results of shaking table tests for −1.0 + 0.50 mm
the teeter water rate. Fe content of the TBS underflow and −0.50 + 0.10 mm size fractions.
increases with an increasing teeter water rate. The con- Coarse size fraction (−1.00+0.50 mm)
centration of Fe was upgraded from the feed value of Weight (%) Fe (%) Recovery (%)
Feed 100.00 56.95 100.00
19.65% to 31.81% with a lower teeter water rate (4 lpm) Concentrate 1 48.60 67.81 57.87
and 49.16% with a higher teeter water rate (7 lpm). It is Concentrate 2 25.37 64.00 28.51
Middling 15.18 44.59 11.88
evident that an increase in the teeter water rate led to a Tail 10.85 9.09 1.73
decrease in the weighted amount and recovery of Fe from Concentrate (Combined) 73.97 66.50 86.38
the underflow. The amount of the underflow stream Fine size fraction (−0.50 + 0.10 mm)
decreased from 55.54% to 30.83%. This could be due to Weight (%) Fe (%) Recovery (%)
Feed 100.00 31.81 100.00
the change in the upward hydraulic transport force. Concentrate 1 21.32 66.97 44.90
Concentrate 2 9.07 62.99 17.96
Increase in teeter water increases the upward water velo- Middling 26.24 35.72 29.47
city in the column towards the overflow stream that in Tail 43.37 5.63 7.67
Concentrate (Combined) 30.39 65.78 62.85
turn pushes the coarser, denser, and locked particles to
the overflow stream. It is observed that the better product
quality of the underflow stream was achieved at a higher
level of teeter water. It can be noted that at a lower teeter a positive effect on performance. Some of the finer iron
water flow rate, there was a lower beneficiation in the minerals may have reported to middling and tail streams
underflow product of fine fraction. An increase in the because of a high wash water rate and tilt angle, but
teeter water could increase the superficial water velocity heavier iron minerals in the coarse size fraction could
in the vertical direction at the separation chamber, which easily be affected by table motion and reported to the
might have carried coarser gangue particles and locked concentrate. The feed grade of the fine fraction was also
particles into the overflow stream. much lower than that of the coarse fraction. This played a
Fe content of all overflow samples was quite low. big role in determining the performance as well. As for the
This can be explained with the positive effect of a middling, re-grind or middling beneficiation options can
narrow size fraction to the beneficiation operations. A be considered because of the particularly high grade of the
narrow feed size distribution clearly minimizes the coarse fraction. It is well known from the liberation stu-
effect of particle size for beneficiation.[16] dies that liberation degree was quite high at each size
The main aim of the rougher circuit was to obtain fraction. A microscopic analysis was performed to the
the highest recovery and produce a clean tail. middlings and it was found that the amount of liberated
Therefore, as the highest Fe recovery was obtained in hematite particles was quite high for both size fractions
the fine fraction with the low teeter water rate, this (Fig. 7). Scale size for coarse fraction was 2000 µm and
sample was decided to be used for the following tests. scale size for fine fraction was 1000 µm. In order for these
liberated hematite particles to beneficiate, middlings were
recycled to table feed.
Gravity concentration studies
The underflow fraction of the TBS contains 56.95% Fe
Wet high intensity magnetic separation studies
content for a coarse fraction and 31.81% Fe for a fine
fraction. Because Fe contents of these concentrates were Another beneficiation route utilizing the differences in
insufficient for making sinter/pellet concentrate, both size and magnetic susceptibility of minerals was con-
fractions were subjected to a shaking table beneficiation ducted. A number of tests were performed by using a
separately to improve the quality. The results of the laboratory scale WHIMS in which different magnetic
tests for both size fractions are given in Table 4. field intensities were applied. The results of the test are
For both size fractions, each concentrate had a satis- given in Table 5.
factory content of Fe, thus concentrates were combined. As seen from Table 5, 73.68% of the feed material from
From the coarse fraction, a combined concentrate with the coarse size fraction was collected as a concentrate with
~74% by weight, 66.50% of Fe grade and 86.38% recovery a Fe grade of 65.18% and a Fe recovery of 84.34% at the
was obtained. Although the combined concentrate’s grade lowest field intensity. In the same field intensity, only
for fine fraction was good, the recovery and particularly 35.96% of feed material was collected as a concentrate
the weight amount, was much less than the coarse frac- with a Fe grade of 65.67% and a Fe recovery of 74.25% at
tion. It can be said that the performance of the coarse size the fine size fraction. The recovery difference between
fraction was better than fine size fraction. Besides that; coarse and fine size fractions can be explained due to
narrow sizes obtained in the coarse fraction may have also the feed Fe contents of the fractions. As a result, combined
2850 O. OZCAN AND I. B. CELIK
Figure 7. Stereo microscopic view of middling products (Top: Coarse fraction, bottom: Fine fraction).
Table 6. The results of slime beneficiation test. gravity and/or magnetic separation techniques. After
Magnetic field Stream Weight, % Fe, % Fe Recovery, % mass balance studies, a dry and two wet notional flow
intensity
(gauss) Feed 100.00 16.68 100.00
sheets have been considered. The weight, grade, and
6960 Concentrate 10.00 66.67 39.96 recovery of all streams were calculated. The dry option
Tail 90.00 11.13 60.04 consists of size separation and high intensity dry magnetic
8625 Concentrate 17.01 52.34 53.37
Tail 82.99 9.37 46.63 separation. The first wet option consists of a TBS as a
9630 Concentrate 22.44 44.79 60.25 rougher beneficiation followed by gravity concentration
Tail 77.56 8.55 39.75
for cleaning while in the second wet option the iron ore
tail concentrated in a similar way but the gravity concen-
tration was replaced by wet high intensity magnetic
recovery (39.96%). As a result, the concentrate obtained at a separation. The slime fraction could not be concentrated
magnetic field of 6960 gauss can be used for sinter/pellet by the TBS because of fine and wide size distribution.
making but, the WHIMS beneficiation of this fraction adds
to only 0.71 to the total mass yield at this grade. Therefore,
slime beneficiation circuit is added as an optional part in the Option 1: Flow sheet including dry screening and a
flowsheets. The best results of each beneficiation study were dry high intensity magnetic Separator
used for flowsheet development.
This flow sheet includes a two-stage dry size classifica-
tion, and a one-stage high intensity dry magnetic
separation (Fig. 8).
Flow sheet options
It can be seen in Fig. 8 that a high grade coarse
It is evident from the beneficiation studies of the iron ore concentrate with 66.40% Fe could be obtained, but if
tail that a sinter/pellet grade product can be produced by it is mixed with fine concentrate then the Fe grade of
the final concentrate can be only 53.05% Fe with a Option 3: Flow sheet including a TBS followed by
recovery of 51.36% Fe. It is explicit that the dry WHIMS
option does not serve for producing a sinter/pellet
This flow sheet differs from Option 2 by the addition of
grade concentrate.
high intensity magnetic separators instead of shaking
tables (Fig. 10).
Approximately 26% of the material can be taken as
Option 2: Flow sheet including TBS followed by the final concentrate with a 65.34% Fe content and the
shaking table total Fe recovery of 70.91%. The final tail includes
9.47% Fe with a 29.09% loss. Slime beneficiation is
This option includes two stages of size classification
added as an optional part in the flow sheet. Slime
followed by two stages of gravity separations by using
fraction (−0.1 mm) can be taken as tail. The impurities
a TBS and shaking table (Fig. 9). Beneficiation of the
that have importance for the sinter/pellet concentrates
slime fraction was also assessed.
were also checked on the final products obtained from
With the TBS followed by the gravity concentra-
Option 2 and 3 (Table 7).
tion option, approximately 30% of the material can
As noticed from the results in flow sheet 2 that the
be taken as the final concentrate with 65.54% Fe
iron grade could be improved to 65.54% with signifi-
content and the total Fe recovery of 68.52%. The
cant low alumina and silica values as 1.50% and 4.96%,
final tail includes 5.79% Fe with 15.56% loss. The
respectively. The overall yield is 30.58% and the overall
middling products were recycled to table feed
recovery of the total iron is 68.52% with respect to the
stream. If an optional slime beneficiation part is
original feed under these conditions.
added to the circuit, only 0.71% concentrate can be
A high grade concentrate composed of a 65.34%
obtained with 66.67% Fe content. In this condition
iron grade with 1.00% alumina and 3.70% silica can
total recovery can be increased from 68.52% to
be produced using flow sheet 3. The overall yield is
70.48%. This part should be considered with operat-
26.11% and the overall total recovery of the iron is
ing costs.
SEPARATION SCIENCE AND TECHNOLOGY 2853
70.91% with respect to the original feed for this ore specifications.[13,14,21–23]. SiO2 content of the
condition. head sample was measured in high amounts such as
It is calculated that 96.64% and 94.81% of silica 45.15 and it could be reduced down to 4.96%. In
and alumina, respectively, were rejected in flow sheet industry the general acceptance limit of SiO2 for
2. Concentrate produced from flow sheet 2 was sui- sinter/pellet feed is given as in the range of 1%–5%
table for the pellet/sinter making (65.54% Fe, 4.96% in these references. It is also reported that SiO2 con-
SiO2 and 1.50% Al2O3) according to different iron tent can be as high as 9.5% in some cases.[22]
2854 O. OZCAN AND I. B. CELIK
The rejection value of silica and alumina in flow Finally, it is evident that upgradation of a low grade
sheet 3 are 97.86% and 97.04%, respectively. Similarly, iron tail can be possible in different beneficiation alter-
concentrate produced from flow sheet 3 was suitable natives using a TBS. It should be noted that if the
for the pellet/sinter making (65.34% Fe, 3.70% SiO2 and proposed flow sheets are considered for operation,
1.00% Al2O3) according to different specifications. they should be re-evaluated in terms of the number of
required equipment, amount of water, and some other
operational parameters considering market prices.
Conclusions
Possibilities for beneficiation of a low grade iron ore
tailing from dry magnetic separation process of an References
existing plant were evaluated in this study. The follow-
[1] Raghukumar, C.; Tripathy, S.K.; Mohanan, S. (2012)
ing conclusions can be deduced: Beneficiation of Indian high alumina iron ore fines – a
case study. International Journal of Mining Engineering
● According to the characterization study; the main and Mineral Processing, 1 (2): 94.
iron-bearing mineral was hematite and it was [2] Jena, S.K.; Sahoo, H.; Rath, S.S.; Rao, D.S.; Das, S.K.;
Das, B. (2015) Characterization and processing of iron
found that liberation size was below 1 mm. But,
ore slimes for recovery of iron values. Mineral
to increase the efficiency of the beneficiation pro- Processing & Extractive Metallurgy Review, 36 (3): 174.
cesses, the feed material was classified as −1.00 + [3] Mohanty, S.; Das, B. (2010) Optimization studies of
0.50, −0.50+0.10 and −00.10 mm. hydrocyclone for beneficiation of iron ore slimes.
● As the existing plant uses dry magnetic separation Mineral Processing & Extractive Metallurgy Review, 31
process, firstly a dry beneficiation route, including (2): 86.
[4] Seifelnassr, A.A.S.; Moslim, E.M.; Abouzeid, A.M. (2013)
classifying the material followed by high intensity Beneficiation of a Sudanese low-grade iron ore.
dry magnetic separation, was evaluated. The com- International Journal of Mineral Processing, 122: 59.
bined final product only gave a 53% Fe grade with [5] Osinubi, K.J.; Yohanna, P.; Eberemu, A.O. (2015)
a recovery of 51.36%. Cement modification of tropical black clay using iron
● Application of the TBS was assessed with two dif- ore tailings as admixture. Transportation Geotechnics,
5: 35.
ferent alternatives including gravity and magnetic
[6] da Silva, F.L.; Araújo, F.G.S.; Teixeira, M.P.; Gomes, R.
separations. Since the unit processing capacity of C.; von Krüger, F.L. (2014) Study of the recovery and
TBS is larger than that of shaking tables and mag- recycling of tailings from the beneficiation of iron ore
netic separators, the TBS reduced the amount of the for the production of ceramic. Ceramics International,
feed to the cleaner stage and increased grade. As the 40: 16085.
Fe contents of the TBS’s overflow were much less, a [7] Chao, L.; Henghu, S.; Jing, B.; Longtu, L. (2010)
Innovative methodology for comprehensive utilization
significant upgrade was achieved just before the of iron ore tailings Part 1. The recovery of iron from
gravity and magnetic separation stages. iron ore tailings using magnetic separation after mag-
● Option 2, in which the TBS is followed by gravity netizing roasting. Journal of Hazardous Materials, 174:
concentration with shaking tables, resulted in very 71.
good quality concentrates. It should be noted that [8] Singh, S.; Sahoo, H.; Rath, S.S.; Sahu, A.K.; Das, B.
(2015) Recovery of iron minerals from Indian iron
the shaking table circuits use large amounts of
ore slimes using colloidal magnetic coating. Powder
water, but this circuit has an advantage with the Technology, 269: 38.
feed fractions being relatively coarse and narrow, [9] Liu, S.; Zhao, Y.; Wang, W.; Wen, S. (2014)
simplifying the dewatering process. Beneficiation of a low-grade, hematite-magnetite ore
● Option 3, in which the TBS is followed by high in China. Minerals & Metallurgical Processing, 31 (2):
intensity wet magnetic separators, also resulted in 136.
[10] Kumar, R.; Srinivas, R.D.; Ram, R.P.S. (2013) Magnetic
a very acceptable quality of concentrate. In this separation studies for a low grade siliceous iron ore
option, both the grade and recovery of final tail sample. International Journal of Mining Science and
were higher than the values from Option 2. Technology, 23: 1.
● The beneficiation of the slime fraction (−0.10 mm [11] Rath, R.K.; Mohanty, S.; Nayak, B.; Singh, R.;
fraction) was also studied. This condition was Bhattacharyya, K.K. (2013) A comparative study on
processing of high alumina hematite iron ore by grav-
integrated as an optional in both flow sheets.
ity, magnetic and flotation methods. Journal of
● Impurity rejection values for the gangue contents Materials Science and Engineering A, 3 (5): 349.
(silica and alumina) in these two flow sheets were [12] Das, B.; Prakash, S.; Das, S.K.; Reddy, P.S.R. (2007)
higher than 94%. Effective beneficiation of low grade iron ore through
SEPARATION SCIENCE AND TECHNOLOGY 2855
jigging operation. Journal of Minerals & Materials [18] Svoboda, J.; Fujita, T. (2003) Recent developments in
Characterization & Engineering, 7 (1): 27. magnetic methods of material separation. Minerals
[13] Shobhana, D.; Santosh, P.; Mohanta, M.K.; Singh, R. Engineering, 16 (9): 785.
(2012) Utilization of iron ore slimes: a future prospec- [19] Das, A.; Sarkar, B.; Mehraotra, S.P. (2009) Prediction of
tive. Separation Science and Technology, 47 (5): 769. separation performance of Floatex Density Separator
[14] Roy, S.; Das, A.; Mohanty, M.K. (2007) Feasibility of for processing of fine coal particles. International
producing pellet grade concentrate by beneficiation of Journal of Mineral Processing, 91 (1–2): 41.
iron ore slime in India. Separation Science and [20] Bhaskar, K.; Govindarajan, B.; Barnwal, J.P.; Rao, K.K.;
Technology, 42 (14): 3271. Gupta, B.K.; Rao, T.C. (2005) Classification studies of
[15] Srivastava, U.; Komar Kawatra, S. (2009) Strategies for lead–zinc ore fines using water-injection cyclone.
processing low-grade iron ore minerals. Mineral International Journal of Mineral Processing, 77: 80.
Processing & Extractive Metallurgy Review, 30 (4): 361. [21] Iron Ore Products. (2009) website; http://vale.basecent.
[16] Tripathy, S.K.; Bhoja, S.K.; Kumar, C.R.; Suresh, N. com/admin/resources/images/2/4/1342.pdf
(2015) A short review on hydraulic classification and [22] Harman, J.; Anglo American Iron Ore Quality and
its development in mineral industry. Powder Volume. (2012) website; http://www.angloamerican.
Technology, 270: 205. com /~/media/Files/A/Anglo-American-PLC-V2/pre
[17] Singh, R.; Maulik, S.C.; Chakravorty, N. (April, 1994) sentations/2012pres/global_iron_ore.pdf
Advances in gravity beneficiation of ores and minerals. [23] Mark Wren, M.; Henry, D.; Iron ore 64.5 % Fe speci-
In: National Seminar on Indian Mineral Processing fication: website; http://www.ipccaustralia.com /down
Advances, Puri. loads/ipcc-iron-ore-64-specifications-brazil-web.pdf