Ottavio Raul Domenico Riberti et al.
Metallurgy andMetalurgia
materials
e materiais
Comparison between bentonite
and serpentinite in the
production process of iron ore
pellets
Comparação entre bentonita e serpentinito no
processo de produção de pelotas de minério de
ferro.
Ottavio Raul Domenico Riberti
Carmignano
Director of Companhia Pedras Congonhas Ltda.
Graduate student attending a M.Sc. course in
Social Economic and Environmental Sustainability
at the Ouro Preto Federal University, State of
Minas Gerais.
E-mail: ottavio@pedrascongonhas.com.br
Abstract
Cornélio de Freitas Carvalho
Professor Associado of the Chemical Department,
Ouro Preto Federal University (UFOP), University
Campus 35.400-000 Ouro Preto, MG, Brazil
E-mail: cornelio@iceb.ufop.br
Pelletizing iron ore fines is an agglomeration process that through a thermal treatment
converts the ultra-fines fraction thereof into small balls ranging in size from 8mm
(0.31 in.) to 18mm (0.71 in.), with adequate characteristics for feeding steel reduction
works. The binder more used to make pellets is bentonite, which is an item of significant cost in the process. The present paper aims at evaluating the use of serpentinite
instead of bentonite. The results obtained show that the full substitution of bentonite
for serpentinite is unfeasible. However a potential does exist for using serpentinite
and bentonite together in the iron ore palletizing process in the proportion of 1:1.
Keywords: iron ore pelletizing, serpentinite, bentonite.
Resumo
A pelotização dos finos de minério de ferro é um processo de aglomeração, que,
através de um tratamento térmico, converte a fração ultrafina em esferas de tamanhos
na faixa de 8mm ( 0,31 pol. ) a 18 mm ( 0,71 pol. ), possuindo características apropriadas para alimentação das unidades de redução das usinas siderúrgicas. O ligante mais
utilizado para a produção de pelotas é a bentonita, que é um item de custo significativo para o processo. Esse trabalho propõe uma avaliação do emprego do serpentinito
em substituição a bentonita. Os resultados obtidos indicam que à substituição total
da bentonita pelo serpentinito é inviável. Mas existe potencial para a utilização do
serpentinito combinado com a bentonita, na razão 1:1, no processo de pelotização de
minério de ferro.
Palavras chave : Pelotização do minério de ferro, serpentinito, bentonita.
REM: R. Esc. Minas, Ouro Preto, 67(1), 47-53, jan. mar. | 2014
47
Comparison between bentonite and serpentinite in the production process of iron ore pellets
1. Introduction
Serpentinite is a rock consisting almost wholly of minerals of the accessory
group of serpentines, such as the antigorite, dolomite, magnetite and magnesite
(NEWMANN and OLIVEIRA, 2003;
BRANDÃO, 2007); it includes lizardite
as one of the minerals of the serpentinite
rock. These minerals are formed by the
serpentinization of igneous rocks, mainly
peridotite (PELLANT, 1992).
The serpentinite may also have
higher silica and lower MgO tenors, and
under such conditions it is employed as an
aggregate in civil construction (SENIOR
ENGENHARIA, 2006).
For the last forty years, it has been
used primarily as a flux (source of MgO)
in the sintering process at integrated
coke-using plants in Brazil (LEMOS et
al., 1979). After the 2008 world economic
crisis, the steel industry started to employ
a metal load of high silica tenor, which
makes the use of serpentinite unfeasible
due to the high silica tenor of said rock.
PIMENTA and his collaborators (2010)
indicate the replacement of serpentinite
by dolomitic limestone as a steel industry
flux from 2010 on.
Other applications of serpentinite
are its use as ornamental stones (ISMAEIL, 2008); its use as aggregates in
civil construction and as a base for asphalt
pavements (RODRIGUEZ, 2007); its
use as a flux in steel plants (WAGNER,
1972); and its use as a raw material for
ceramic products (GUR’EVA, 2009).
Possible uses include the potential treatment of mine acid drainage (BERNIER,
2005); in the carbon dioxide sequestering
(TEIR, 2007); and its use in producing
magnesium oxide (GLADIKOVA, 2007).
The production of iron ore pellets
employing serpentinite as a source of
MgO was already described by FONSECA (2003) and ARAGÃO et al (2000).
Notwithstanding these early experiences
the use of serpentinite has been passed
over in view of the presence of structural
water, which causes a loss due to a greater
calcination in comparison with other
MgO sources. However when serpentinite
is used in the production of ore pellets, the
iron and its other chemical components,
such as silicon oxide (SiO) and iron oxide
(FeO), have not been taken into account,
as well as their application as a clinging
agent, the role of which is kept for the
bentonite, an aluminum silicate that has
various industrial applications and chemical resemblance with serpentinite.
During the palletizing process the
ore is mixed with limestone, the binder
and coal. The limestone functions are to
strengthen the raw and the burnt pellets;
to hinder the rupture of the raw pellets
while drying off inside the furnace; to supply CaO to the mixture to be processed; to
adequate the chemical and metallurgical
features of the fired pellets; to lower the
burning temperature required to make
the acid cangue cling to the ore grains.
The binder functions, usually bentonite,
are to promote and ease the formation
of the raw pellets during the pelletizing
operation; to give the raw pellets the
conditions needed to resist transportation
and handling that follow the unloading
from the palletizing disc up to the indurating furnace; to optimize the dry and wet
strength of the raw pellet; to improve the
strength to the thermal shock during the
drying phase.
The coal functions are adding thermal energy into the process, reducing
fuel oil consumption (that is one of the
more expensive inputs thereof), yielding
gains to the physical quality of the pellet,
for instance, strength to compression,
improvement in the metallurgical quality
due to the increased number of pores in
the pellet.
The present paper aims at carrying
out an evaluation of the action in the
iron ore pellet production, of both the
serpentinite and the bentonite, an aluminum silicate that has many industrial
applications and chemical similarities
with the serpentinite. The binding characteristic of the serpentinite was observed
by JANUZZI(2008).
From the economic viewpoint bentonite has an average price of US$107.00
in the regions it is produced (SILVA et
al. 2008), while serpentinite has in its
production areas an average price of
US$14.90 (CHENG et al., 2002). But
they are different minerals: bentonite is
an aluminum silicate supplied in a fine
grain size specified in meshes; serpentinite
is a magnesium silicate and is supplied
in a larger grain range specified in millimeters. Taking into account the need of
having to be milled, we can affirm that
the processed serpentine price, although a
competitive price, will be higher than the
mentioned one, in view of the difference
between the two mentioned prices.
Also, the site of the deposits should
be considered in view of the need of hauling the ores from their origins up to the
consumption market. And the cost of
freight may be higher than the cost of the
raw material.
where Pedras Congonhas Extração Arte
Indústria Ltd. mines said rock since 1979.
Utilized was the fines fraction that is the
waste generated by the mining of the
serpentinite. The bentonite came from
Empresa União.
The mixes developed were prepared
in an intensive mixer manufactured by EIRICH, model R 02, that has a capacity of
up to 5.0 liters (9.08 dry pints). The pellets
were produced using a tire pelletizer with
a three-phase 0.5 c.v. (0.493 hp) motor.
2. Methodology
For the pelletizing tests the iron ore
samples have come from Vale Company’s
Fábrica Mine. The serpentinite under
study is originated in the Corrego Dos
Boiadeiros Geologic Formation, at Nova
Lima County, State of Minas Gerais,
Table 1
Mixes made for the pelletizing tests
48
REM: R. Esc. Minas, Ouro Preto, 67(1), 47-53, jan. mar. | 2014
Ottavio Raul Domenico Riberti et al.
Three types of pellets were produced, as described below in Table 1.
To perform the tests 200 kg (442
lb) of iron ore were used. The binder was
used at ratios of 6kg/metric ton of ore, 54
kg/metric ton of limestone and 13.5 kg /
metric ton of coal.
For the pellet size distribution tests,
adopted was the Brazilian standard NBR
ISO 2395 – Testing Sieves and Sieving
Tests, and the standard ISO 9045 (industrial sieving), both enacted by ABNT
– Brazilian Technical Standards Association. The sieves utilized comply with
ASTM -11 – 70 specification, trade mark
Bronzinox, and the maker of the mixer is
Produtest. The test consists in sampling
around 1.5 kg (3.31 lb) of pellets during
the production thereof and sieving the
resulting sample in the meshes between
5mm and 18mm, weighting the fractions
obtained and calculating the percentage
retained on each mesh.
For the humidity test the Brazilian
standard NBR 3087 methodology was
used. A stove at 110º Celsius and an
analytical scale were used. The determination of the specific area was made
through the Fischer parameter (Brazilian
standard ABNT NBR 6221).
For the chemical analysis the X-ray
fluorescent technique was used together
with a Rigaku Simultix X-ray spectrometer (using cast lozenges).
The methodologies utilized for
determining the water absorption, and
determining colloids and swelling, have
been the following:
Water absorption – to put a porous
plate within a porcelain capsule suspended about 1 cm above the capsule bottom and resting on aluminum supports
placed at the 4 corners and the center of
the plate. Add distilled water up to the half
of the plate. Get 2.00 grams of the sample
on a rapid filtration paper. Place the paper
with the sample on the porous plate within
the receptacle for absorption. Wait for two
hours. Weigh the sample together with the
paper filter. Make a white test. Calculate
the water absorption percentage.
Determination of colloids – agitate
with a magnetic agitator 10.00 grams of
the sample in 1000 mL for 10 minutes.
Apply centrifuging for 15 minutes at
approximately 1400 rpm to two aliquot
parts of 100 mL. Weigh the floating liquid of one of the aliquot parts in a 250
mL beaker whose weight was previously
determined. Dry completely in a stove at
105ºC. Cool in a desiccator and weigh the
sample. The percentage colloidal tenor is
the sample mass that remains in the beaker
multiplied by 100.
Determination of swelling – add
2.00 grams of the sample, previously
dried off at 105ºC, in fractions of about
0.1 gram in 100mL of distilled water
contained in a test tube. Let the sample
rest for 4 hours; read the sample volume
in the test tube.
For tests of the number of raw pellet
tumbles 15 pellets between 12,5mm and
10.0mm have been randomly picked. Roll
every one of the raw pellets and let them
fall freely and repeatedly from a height of
45cm until the first crack appears. Count
and record how many tumbles each pellet had until the first crack occurred. The
result will be the arithmetical average of
the recorded figures and will be expressed
in number of tumbles per pellet.
For the tests of raw pellets compression strength, 15 pellets will be randomly
picked and a Kratos-trade-mark press is
used, equipped with parallel pans, a 10kg
capacity dynamometer, and a 0.005 kg/
pellet measuring precision. Dry pellets at
105 ± 5ºC should be cooled to the ambient temperature before the test. Record
the load registered in the press scale at
the rupture moment showed by the dynamometer finger. The result will be the
arithmetical average of the recorded values
and is expressed in kgf/pellet. In the case
of the pellet firing the temperature and
the test residence time were respectively
1320ºC and 24 minutes.
As references for the compression
and abrasion strength tests the ABNT
standards NBR 4700 and 6465 were used.
The tests were held in a pilot plant
of a Brazilian company that produces iron
ore pellets.
The chemical composition of the
iron ore used in the pelletizing tests is
shown in table 2. Tables 3 and 4 present
the chemical composition and important
physical-chemical properties of respectively the serpentinite and bentonite.
Table 2
Chemical composition of the iron ore
utilized.
*PPC = Loss on calcination
Table 3
Chemical compositions of serpentinite
and bentonite used in the tests.
REM: R. Esc. Minas, Ouro Preto, 67(1), 47-53, jan. mar. | 2014
49
Comparison between bentonite and serpentinite in the production process of iron ore pellets
Table 4
Physical-chemical properties of serpentinite and bentonite.
As for the bentonite, the higher
the sodium tenor, the greater the water
absorption, and therefore the greater the
swelling is. However, even worse is the
peeling of layers because it requires more
energy for opening the bladed mineral,
that is, a trend to form isolating gel (need
of a good sheering stress capable of overcoming the force among the clay-mineral).
The colloid, in turn, is important
in the palletizing process, as it implies a
greater covering area. The colloid means
particles below 2µm, that is, a greater
m2/gram of agglomerating action. The
ideal is to have a high colloid tenor and
a minimum swelling, but usually if one
is high, the other will also be high. The
water absorption is an indirect measure
of swelling.
The binding materials normally
present high specific surface. The bentonite specific surface determined for this
research bentonite (table 4) is lower than
that of lime, that is a good binder (6558.21
cm2/gram)(Bezerra et al., 2011)
When one compares serpentinite
with bentonite, table 4, one sees that the
water absorption, colloids, swelling, material fineness parameters are unfavorable to
the serpentinite when the agglomeration
process is considered. The same occurs
in relation to the serpentinite low sodium
tenor, table 3, in relation to bentonite.
As to the chemical parameters,
the lesser SiO2 tenor of the serpentinite
is highlighted favorably in relation to
the bentonite’s. But the high MgO tenor
would limit the serpentinite utilization
only in the production of pellets with a
high MgO tenor.
Other parameters that do not favor
the use of serpentinite as a binder is the
predominance of fines (99% passing the
44µm sieve) and its high specific surface
in relation to the bentonite.
Table 5 presents the raw pellets test
results.
The mechanism of forming the
Figure 1 shows the number of pellet
tumbles from a 45 cm (17,72 in.) height.
This test consists in carrying out
consecutive raw pellets tumbles on a steel
surface, from a 45cm height, in order to
determine the average numbers of times
that these pellets resist the tumble until
the first crack appears.
The analysis of Figure 1 shows
that test no. 2, that contains 50% of the
serpentinite, presents a number of 45cm
( 17,72 in. )-tumbles near the tumbles of
test 1, containing only bentonite. In this
case, the palletizing process that uses the
serpentine alone, the reduction of the
number of raw pellets tumbles decreases
as much as reaching 26.5%.
Figure 2 presents the strength (kgf/
pellet) for the raw pellets. This test shows
the power necessary to destroy the raw
pellet. Considering that the raw pellet
should be handled until being transported
to the furnace to be burnt, this information is of vital significance to avoid that
the pellets arrive at the furnace without
mechanical degradation.
The analysis of Figure 2 is similar to
the analysis of Figure 1, i.e., it shows that
test no. 2, that contains 50%w/w of serpentinite, presents a Strength (kgf/pellet)
near the one obtained in test 1, containing
only bentonite, with a strength reduction
Table 5
Tests with raw pellets.
pellets occurs from a phenomenon that
involves a solid phase and a liquid phase,
the former being characterized by the
mixture of iron ore, binding agents and
additive agents, and the latter by water
(MENDES, 2009).
When humidity is increased or reduced, interference will occur in the size
of the pellets and the speed in which they
are formed.
The mixture no. 1 humidity is a typical one when only bentonite is used as a
binder. The humidity of mixture numbers
2 and 3 were reached empirically so as to
produce pellets with the same size of the
pellets produced using only bentonite.
50
REM: R. Esc. Minas, Ouro Preto, 67(1), 47-53, jan. mar. | 2014
Ottavio Raul Domenico Riberti et al.
of 10%, while the no. 3 mixture presents
a 40% strength reduction in comparison
with mixture no. 1.
As to the strength of burnt or dry
pellets Figure 3 presents the tests results
obtained. This test shows the amount
of power necessary for the mechanical
degradation of the fired or dry pellet.
Considering that the fired or dry pellet
should be transported up to a stock pile,
where it shall be store piled, and in the
future loaded into ships, wherein it will
Figure 1
Number of 45cm ( 17,72 in. ) tumbles
of the mixtures as a function of the test
number.
Figure 2
Strength of raw pellets as a function of the
test number.
Figure 3
Dry pellets strength as a function of the
test number.
again be piled, such information is of
vital importance to avoid destruction
of the burnt and dry pellets along these
operational phases.
One sees that the Strength drops as
the bentonite tenor is reduced; the no. 2
test, that has 50% m/m of serpentinite,
presents a loss of 37% in its strength, while
the no. 3 test, that has 100% m/m serpen-
tinite, presents a strength loss of 67%.
Table 6 presents the results of the
burnt pellets abrasion test.
In accordance with OLIVEIRA
(2010), the results in the range of +6.3
mm should be higher or equal to 95%.
Only test no. 3 didn’t reach said figure,
although reaching a very near value
(94.9%). As such, tests 1 and 2 are within
the acceptable limits. For the -0.5mm
range the results should be lower than
5%. Therefore, tests 1 and 2 are within
the acceptance limits.
Table 7 presents the results of the
compression strength tests of burnt pellets.
The compression strength test of
burnt pellets shows which is the burnt pellets percentage that presents a mechanical
REM: R. Esc. Minas, Ouro Preto, 67(1), 47-53, jan. mar. | 2014
51
Comparison between bentonite and serpentinite in the production process of iron ore pellets
Table 6
Burnt pellets abrasion tests.
Table 7
Compression strength tests of burnt
pellets.
wear when submitted to a force lesser than
250 daN and when submitted to a force
lesser than 100 daN. Considering that the
fired or dry pellet should be transported
up to a stock pile, where it shall be store
piled, and in the future loaded into ships,
wherein it will again be piled, such information is of vital importance to avoid
destruction of the burnt and dry pellets in
these operational phases.
The average strength of pellets produced with 50% bentonite and 50% serpentinite is 1.5% lower than the strength
produced only with bentonite, but the
pellet produced only with serpentinite is
14.9% lesser.
Regarding the percentage of pellets destroyed when submitted to a force
lesser than 100 daN, those produced with
50% of serpentinite and 50% of bentonite
showed an 0.8% increase in the quantity
of destroyed pellets, while the use of only
serpentinite increased that quantity by
1.6%.
These results show that there was a
loss of strength of the burnt pellets when-
ever the serpentinite was used, and that
the use of only serpentinite jeopardizes
their strength.
Through the analysis of the full
substitution of bentonite for serpentinite
one sees a light fall in the SiO2 tenor
(from 3.70% to 3.63%) and an increase
of 0.18% in the MgO tenor.
shows that such a combined application
of the two products is feasible, provided
the quality meets the needed required,
a MgO tenor of around 0.20% is acceptable in an end product, and some
process adjustment is carried out seeking to minimize the generation of fines
during the firing phase.
Although the serpentinite doesn’t
present a binding characteristic, due to
its size distribution (99% passing the
325 mesh and high specific surface), it
presents a potentiality to be used in a
combined manner with the bentonite.
The use of serpentinite may re-
duce the cost of pelletizing iron ore,
depending on the location site of this
rock and the bentonite and the particularities process from the pelletizing
producer.
It is recommended to carry out
supplemental tests in order to confirm
the results obtained and check the
porosity and reducibility of the pellets
manufactured and its microstructure.
It is also possible to modify the percentages of serpentinite and bentonite
in the mixtures and work with a more
adequate size distribution in the serpentinite.
4. Conclusion
Taking parameters Na2O – activation level, water and colloids absorption, the serpentinite is not characterized as a binder.
The use of 100% of serpentinite
in the palletizing tests produced raw
pellets with low compression strength.
Also, in the firing phase the abrasion
and compression results showed a worsening, which indicates the impossibility
of the full substitution of bentonite for
serpentinite.
In the fifty-fifty substitution (50%
bentonite/50% serpentinite) an intermediate quality was obtained, which
52
REM: R. Esc. Minas, Ouro Preto, 67(1), 47-53, jan. mar. | 2014
Ottavio Raul Domenico Riberti et al.
Acknowledgments
To João Júlio Tolentino, Thiago Marchezi Doellinger and Vinicius Oliveira Fonseca for assistence and contribution in laboratory tests.
References
ARAGÃO,R. et al. Desenvolvimento de pelotas de minério de ferro com alto teor de MgO na
Samarco Mineração S/A. In SEMINÁRIO DE REDUÇÃO DE MINÉRIO DE FERRO, 31,
2000,Santos,São Paulo,p. 865-878.
BERNIER, L. R. The potential use of serpentinite at the passive treatment of acid mine drainage.
Environ Geol., n. 47, p. 670 - 684, 2005.
BRANDÃO, P. Caracterização da rocha serpentinito. Relatório interno da Pedras Congonhas
Ltda. Escola de Engenharia da Universidade Federal de Minas Gerais – EE/UFMG. Belo Horizonte, December, 2007
CHENG, T.W. et al. A study of synthetic forsterite refractory materials using waste serpentine
cutting. Mineral Engineering, v.15, p. 271-275, 2002.
FONSECA,V. Envelhecimento de pelotas de minério de ferro com diferentes basicidades e teores
de MgO. Universidade Federal de Ouro Preto, 2003. (Ms. Dissertation)
GLADIKOVA, A. et al. Evaluation of reaction variables in the dissolution of serpentine for mineral carbonation. Elsevier, Fuel 86 ( 2007) 273-281.
GUANZHOU Q. et al. Functions and molecular structure of organic binders for iron ore pelletization.Colloids and Surfaces A: Physicochem. Eng. Aspects, n. 224, p. 11-22, 2003.
GUR´EVA, V.A. Application of magnesium-bearing technogenic raw material in the production
of decorative-finishing ceramics. Glass Ceram, v. 66, p. 95-98, 2009.
ISMAIEL, I. S. Characterizations of some Egyptian serpentinite used as ornamental stones. Springer Link (MetaPress), Chin J. Geochem. v. 27, p. 140-149, 2008.
JANUZZI,A. Análise da aglomeração a frio no processo HPS (Hybrid Pelotized Sinter) com
ênfase nas matérias-primas envolvidas. Universidade Federal de Minas Gerais, 2008.(M.Sc.
Dissertation)
LEMOS,M. et al. Avaliação técnica de matérias-primas para sinterização. Metalurgia, ABM, v.
35, no. 257, 1979.
MARTINS.J. et al. Emprego de resíduos siderúrgicos e pellet feed minipelotizados na sinterização
de minério de ferro. REM – Revista Escola de Minas, v. 56, n. 4, p. 249-254, 2003.
MENDES, J. Influência da adição de lama fina de aciaria a oxigênio nas características físicas
e microestruturais de pelotas queimadas de minério de ferro.Universidade Federal de Ouro
Preto, Redemat, 2009. (M.Sc. Dissertation).
MOHAMED,O.A. et al. The role of normal and actived bentonite on the pelletization of barite
iron ore concentrate and the quality of pellets. Powder Technology, v. 130, p. 277-282, 2003.
NEUMANN, R. et al. Caracterização mineralógica e tecnológica de minério de serpentina. RT
2003-011-00. Centro de Tecnologia Mineral – CETEM. Rio de Janeiro, February, 2003 . 18p.
OLIVEIRA, V. Estudo da porosidade de pelotas de minério de ferro para altos-fornos através de
adsorção física. Universidade Federal de Minas Gerais, 2010. (M.Sc. Dissertation)
PELLANT,C. Rocks and Minerals. Londres: Dorling Kindersley Book,1992 . P. 194.
PIMENTA,F. et al . Elevação do teor de MgO do sínter com o aumento da utilização de calcário
dolomítico. SEMINÁRIO DE REDUÇÃO DE MINÉRIO DE FERRO, 40,Belo Horizonte:
2010.
PINTO, P. Uso do esteatito como componente de espumas de poliuretano flexível.Universidade
Federal de Minas Gerais, Instituto de Ciências Exatas, Departamento de Química, 2011.
SENIOR ENGENHARIA. Reavaliação de reavaliação de reservas minerais. Town of Nova
Lima, Minas Gerais, April, 2006.
SILVA, A.R. et al. Bentoniticclays : concepts, structures, properties, industrial uses, reserves, production and national and international producers/suppliers. Revista Eletronica de Materiais e
Processos, v.32, p. 26-35, 2008.
TEIR, S. et al. Production of magnesium carbonates from serpentine for long-term storage of
CO2. Elsevier, International Journal of Mineral Processing, v.85, p.1-15, 2007.
WAGNER, E. et al. Pre-Hispanic Workshop of Serpentinite Artifacts, Venezuelan Andes, And
Possible Raw Material Source. Sciense, V. 175, p. 888-890, 1972.
Artigo recebido em 11 de janeiro de 2013. Aprovado em 13 de setembro de 2013.
REM: R. Esc. Minas, Ouro Preto, 67(1), 47-53, jan. mar. | 2014
53