Effect of iron ore-coal pellets during reduction with hydrogen and carbon
monoxide

Yi MAN, Jun-xiao FENG

PII: S0032-5910(16)30449-1
DOI: doi: 10.1016/j.powtec.2016.07.057
Reference: PTEC 11819

To appear in: Powder Technology

Received date: 29 November 2015

Revised date: 21 July 2016
Accepted date: 22 July 2016

Please cite this article as: Yi MAN, Jun-xiao FENG, Effect of iron ore-coal pellets
during reduction with hydrogen and carbon monoxide, Powder Technology (2016), doi:
10.1016/j.powtec.2016.07.057

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Effect of iron ore-coal pellets during reduction with

hydrogen and carbon monoxide
Yi MANa,b*, Jun-xiao FENGc

T
a Nanjing Vocational Institute of Industry Technology, Jiangsu 210000, China

b School of Material and Metallurgy, University of Science and Technology Liaoning, Liaoning 114000, China

IP
c School of Mechanical Engineering, University of Science and Technology Beijing, Beijing 100083, China

Abstract: In this study, the effects of process parameters, such as gas composition and reduction temperature,

R
were investigated. X-ray diffraction and scanning electron microscope were used to identify the products in

SC
different atmospheres. Experimental results indicate that the three-dimensional diffusion controlled model is suited
to the reduction process. The compressive strength of the produced pellets gradually decreased as the temperature
increased from 700°C to 900°C in all atmospheres. The formation of the iron whiskers made the adjacent grains

NU
loose and eventually caused the decrease of the compressive strength. The presence of H2 atmosphere had a strong
effect on the pellets reduction due to the formation of porous and fibrous iron.
MA
Key words: Iron ore-coal pellets; Direct reduction; Reduction kinetics; Reduction characteristics

1. Introduction
Direct reduction of iron ore (DRI) has become an important new process for iron making,
D

which is due to the importance of iron and steel in the current and future technologies [1]. The
TE

reduction of iron oxide is a gas–solid reaction, which has been extensively studied by numerous
investigators in the past several decades. Rao [2] investigated the reduction kinetics of hematite
and carbon powders in the temperature range of 850～1087°C. A mathematical time-dependent
P

and isothermal model based on the grain model has been developed to simulate the kinetic and
CE

thermal behaviors of a porous iron oxide pellet undergoing chemical reactions with a mixture of
hydrogen, carbon monoxide, carbon dioxide and water vapor by Valipour et al. [3]. The reduction
of Fe2O3 by H2 in the temperature range of 220～680°C and the reduction of Fe3O4 by H2 in the
AC

temperature range of 210～950°C were discussed by Pineau et al. [1, 4]. Halder et al. [5-7]
examined the rates and shrinkage of Iron-Oxide-Carbon Composites. The reduction kinetics of
Fe2O3 to FeO in the presence of syngas constituents was investigated by Piotrowski [8] based on
the thermogravimetric data. Weiss et al. [9] investigated the reduction kinetics of iron ore fines
under fluidised bed conditions in the temperature range of 400～700°C. The effects of briquettes
composition, reduction temperature and reducing gas composition on the reduction of biomass and
iron oxide were discussed by Luo et al. [10]. Iron ore reduction by the gaseous products of
methane partial oxidation in a porous media (H2 and CO) was studied numerically and
experimentally by Palacios et al. [11]. The influences of temperature and time on the reduction
rate of iron ore pellets with and without biomass was studied using hydrogen as the reducing agent
by Guo et al. [12]. Currently, the research of direct reduction has mostly focused on energy saving,
reducing consumption, reduction characteristics and reaction rate. Due to the shortage of natural
gas in China, coal-based direct reduction still occupies an important position. Serious problems
encountered in the process of Coal-based direct reduction are the sticking and unascertained

*
Corresponding author. Tel.: + 86 13720027590; fax: + 86 10 62332741
E-mail address: bestmanyi@sina.com (Y. Man)
 ACCEPTED MANUSCRIPT

reduction mechanisms. With the development of new reduction technology, the study of reduction
mechanisms and sticking phenomenon are becoming more and more urgent. Yi et al. [13]
examined the volume change of pellet reduced in H2-CO mixtures at 800-1000°C with a novel
consecutive imaging method. The phenomenon of sticking was recognized by many former
investigations on the reduction of iron ore [14-18]. The sticking behavior of iron ore–coal pellets

T
(ICPs) at high temperatures under a load of 0.1 MPa was studied by Li et al. [19]. However, the
research on the comparison of the models and the change of morphology in different atmospheres

IP
is relatively small, similar studies on the sticking behavior of iron ore-coal pellets in reducing
atmosphere have not been reported.

R
In this research, the effect of temperature and gas composition on reduction mechanism and

SC
the change of morphology in iron ore-coal composite pellets are investigated. The model
equations are applied to estimate the reduction mechanism in different atmospheres. The
morphology is identified by compressive strength, X-ray diffraction and scanning electron

NU
microscope.
2. Materials and experiments
2.1. Materials
MA
The iron ore and coal used in this work were obtained from Hebei and Shanxi province in
China, respectively. The chemical composition of iron ore are shown in Table 1. The average
particle size of iron ore was 74μm. The main chemical composition of coal were fixed carbon =
74.48%, volatile matter =12.51%, and ash = 13.01%. The particle size ranged from 74 to 147 μm.
D

The coal was devolatilized at 600 °C for 1 hour in nitrogen before experiment.
TE

Table 1
2.2. Experimental
A mixture of fixed C (CFix) and oxygen O (OTFe) in iron ore-coal particles at a mole ratio of
P

0.5:1 in reducing atmosphere (CO or H2) were used throughout the study. Pellets of size 15 mm
CE

were made from 1.5% bentonite binder and iron ore–coal mixture in a disc pelletizer
(diameter=1000 mm, rotating=24 rpm, angle of inclination=45°). The pellets were dried in an
oven at 110°C for 2 h before experiment, and these pellets were isothermally reduced at
AC

700-1100°C under reducing atmosphere with a flow of 0.5L/min.

All experiments were conducted in a thermogravimetric analyzer apparatus. This apparatus
comprised a gas distributor disk, a temperature controller, a quartz glass tube reactor, and an
electronic balance. A vertical resistance furnace with a maximum working temperature of 1250 °C
was used for heating. At the beginning of the experiment, nitrogen was passed to prevent
oxidizing atmosphere, and then the experiment was maintained under reducing atmosphere. When
the temperature was raised to the required temperature and maintained, the pellets were placed in
hot zone. The mass of the sample was continuously recorded every 1 min. At the end of the
experiment, the sample was removed from the furnace and cooled in N2 atmosphere.
The reduction fraction  was defined as the following formula [20]:


 m0  mt  100 (1)
O  mass  +C (mass)
where α is the degree of reduction (%), m0 is the initial mass of the sample after moisture
removal, and mt is the mass of the sample after each time t. C (mass) is calculated for the total
mass of fixed carbon, and O (mass) is calculated for the total mass of oxygen presented in each
 ACCEPTED MANUSCRIPT

mixture in the forms of Fe2O3, Fe3O4, and FeO.

3. Results and discussion
3.1. Effect of temperature on the quality of the produced pellets
Considering that the carbon gasification reaction is slow when the temperature is low, the
effect of temperature on reduction reaction was investigated under isothermal conditions

T
(900°C-1100°C) in different atmospheres and the results are illustrated in Fig.1.
Fig.1.

IP
It shows that the rise of temperature led to an increase of reduction fraction for all
atmospheres. The reduction reaction showed a similar trend in CO atmosphere and H2 atmosphere

R
at 900°C. However, the carbon gasification reaction was inhibited by the presence of carbon

SC
monoxide, and the reaction fraction in H2 atmosphere was higher than in CO atmosphere. The rate
of reduction increased obviously when the temperature rose to 1000 °C and the rate of reduction
reactions was significantly more rapid in H2 atmosphere. At 1100 °C, the curve was very similar

NU
when the pellets were reduced to 65%, which could be due to the rapid carbon gasification
reaction.
3.2 Kinetics model and activation energy values
MA
The kinetics equation can be applied to understand the reduction mechanism of iron ore–coal
pellets in different atmospheres.
The iron ore-coal pellets were reduced by the following reactions in N2 atmosphere:
Direct Reductions C+FexOy=FexOy-1+CO （2）
D

Indirect reductions CO+FexOy=FexOy-1+CO2 （3）

TE

Gasification of coal C + CO2 = 2CO （4）

Since the rate of the indirect reduction reactions (3) are much faster than the direct reduction
reactions (2) in the presence of CO, the reduction of iron ore is considered mainly composed of
P

indirect reduction and carbon gasification. Fixed carbon is considered to generate the carbon
CE

monoxide. The limiting step may not be gasification of coal in CO atmosphere and H2 atmosphere.
In order to find the limiting step of reduction reaction, the equations of three-dimensional
phase-boundary controlled reaction and three-dimensional diffusion controlled reaction were
AC

selected to contrast the reduction process, as shown in table 2.

Table 2
Where t is time (s) and k is the rate constant (s-1). Applying the regression method, the values
of kinetics model were calculated. The mean deviation (MD) and root mean square deviation
(RMSD) were selected for comparing the kinetics model and experimental data. The following
equations were applied[8]:

MD 
 cal   exp
(7)
n

    exp 
2

RMSD 
cal
(8)
n 1
Where αcal is the values of kinetics model and αexp is the experimental data. The calculated
results are listed in table 3.
Table 3
By comparing the calculated results in CO and H2 atmosphere, it can be concluded that
 ACCEPTED MANUSCRIPT

three-dimensional diffusion controlled reaction model is best suited to model the experimental
data. Three-dimensional phase-boundary controlled reaction model was evaluated as nonlinear in
the entire reduction process, which suggests that the resistance was small. However, the calculated
error of three-dimensional diffusion controlled reaction model was larger in H2 atmosphere than in
CO atmosphere. Comparing both MD and RMSD values, it can be found that the diffusion

T
resistance decreased in H2 atmosphere. The rate of gas diffusion has been effectively improved.
The model of three-dimensional diffusion controlled reaction was selected and the apparent

IP
activation energy was calculated based on the Arrhenius equation as follows:
E

R
ln k  ln A  (9)
RT

SC
Where E is the activation energy (J/mol), R is the gas constant (J/mol K), A is the frequency
factor (s-1), and T is the temperature (K). The plots of ln (k) versus 1/T are presented in Fig. 2, and
the slopes are then used to estimate the activation energy values.

NU
Fig.2.
From the slopes of these straight lines, it is found that the activation energy of the reduction
reactions were 149.21 kJ/mole in CO atmosphere and 148.82 kJ/mole in H2 atmosphere,
MA
respectively. 3.3. Effect of gas composition on the quality of the produced pellets
The effect of gas composition on compressive strength of the produced pellets was
investigated and the results are illustrated in Fig. 3.
Fig.3.
D

Initially, the compressive strength of the produced pellets gradually decreased as the
TE

temperature increased from 700°C to 900°C in all atmospheres, and showed similar values.
However, the produced pellets had a very large increase in the compressive strength above 900°C.
It is found that with the temperature increased from 900°C to 1100°C, the compressive strength of
P

the produced pellets increased in all atmospheres and reached to their maximum values at 416.2N
CE

in N2 atmosphere and 261.7N in CO atmosphere, respectively. For the produced pellets in H2

atmosphere, this value was 193.2N, which was considerably lower than the produced pellets in
other atmospheres. The presence of H2 atmosphere could lead to the decrease of the compressive
AC

strength at different extent. The increase of the compressive strength may be due to the more
serious sticking behavior in N2 and CO atmosphere.
Fig.4.
X-ray diffraction (XRD) with Cu-K radiation was used to identify the formed phases of the
reduced pellets in different reducing atmosphere as shown in Fig.4. At 900°C, the major phases
were wüstite (FeO) and metallic iron (Fe) in CO atmosphere. However, the mass percentage of
wüstite (FeO) decreased rapidly in H2 atmosphere. It can be concluded that the minimum
compressive strength of the pellets appeared as the formation of plenty of ferrous oxide and iron
whiskers. At 1000°C, the metallic iron was the major phase in reduced pellets, followed by a small
amount of wüstite (FeO). The metallic iron increased with the rise of temperature and eventually
dominated the sample at 1100°C.
To get detailed knowledge about the relation between the quality of the produced pellets and
gas composition, the morphology of produced pellets in different atmospheres was analyzed by
scanning electron microscope.
Fig.5.
As can be seen in Fig. 5 (a, d), the reduced pellets became porous and the products became
 ACCEPTED MANUSCRIPT

fibrous shape in atmosphere of both CO and H2 atmosphere. The formation of iron whiskers made
the adjacent grains loose and eventually caused the decrease of the compressive strength at 900℃.
This kind of sticking was initiated by the contact of the needles that hooked mechanically the
grains together and mostly appeared at the lower temperatures [19]. At 1000 °C, the particles
exhibited a more developed fibrous shape as shown in Fig. 4 (b, e). In addition, it is observed that

T
some agglomeration particles appeared which directly caused the obvious increase of the
compressive strength. The diffusion and crystallization of the newly formed iron improved with

IP
the rise of temperature. Fig. 5 (c, f) depicts the microscopic structure of the reduced samples
obtained from the isothermal experiment conducted at 1100°C. It shows that the iron precipitation

R
differed widely in CO and H2 atmosphere at 1100°C. The precipitated iron was dense and

SC
agglomerated in the atmosphere of CO. However, the precipitated iron was porous in the presence
of H2. It is observed that the iron precipitation in H2 atmosphere exhibited quantity of tiny pores
and more fibrous than in CO atmosphere. The presence of H2 atmosphere had a strong influence

NU
on the reduction of the pellets. The increase of reduction rate in H2 atmosphere could be due to the
increased number of pores and the higher reducibility of H2.
4. Conclusions
MA
The influence of atmosphere, temperature and compressive strength on the reduction
behavior of iron ore-coal pellets was investigated. Kinetics parameters were estimated from
thermogravimetric data. X-ray diffraction and scanning electron microscope were used to identify
the formed phase and the morphology of reduced pellets. The following conclusions can be drawn
D

from this work:

TE

1. The reduction reaction showed a similar trend in CO atmosphere and H2 atmosphere at

900°C.The presence of carbon monoxide inhibited the carbon gasification reaction, and the
reaction fraction in the presence of H2 was higher than in CO atmosphere at lower temperature.
P

However, the carbon gasification reaction became rapid at 1100°C, and played a significant role at
CE

the initial period of reduction.

2. The three-dimensional diffusion controlled model is suited to the reduction process. The
calculated error of the three-dimensional diffusion controlled reaction model was larger in H2
AC

atmosphere, indicating that the diffusion resistance was lower in H2 atmosphere than in CO
atmosphere.
3. The compressive strength of the produced pellets gradually decreased as the temperature
increased from 700°C to 900°C in all atmospheres, and showed similar values. The tendency
shifted as the reaction temperature increased. The compressive strength of the produced pellets
increased in all atmospheres as the temperature increased from 900°C to 1100°C. The minimum
compressive strength of the pellets appeared as the formation of plenty of ferrous oxide and iron
whiskers at 900°C. The formation of iron whiskers made the adjacent grains loose and eventually
caused the decrease of the compressive strength.
4. The iron precipitation differed widely in CO and H2 atmosphere at 1100°C. The
precipitated iron was dense and agglomerated in the atmosphere of CO. However, the precipitated
iron was porous in the presence of H2. The presence of H2 atmosphere had a strong influence on
the reduction of the pellets due to the formation of porous and fibrous iron. The increase of
reduction rate in H2 atmosphere could be due to the increased number of pores and the higher
reducibility of H2.
 ACCEPTED MANUSCRIPT

References
[1] A. Pineau, N. Kanari, I. Gaballah, Kinetics of reduction of iron oxides by H2 Part I: Low
temperature reduction of hematite, Thermochim Acta. 447 (2006) 89-100.
[2] Y.K. Rao, The kinetics of reduction of hematite by carbon, Metallurgical and Materials
Transactions. 2 (1971) 1439-1447.

T
[3] M.S. Valipour, M.Y.M. Hashemi, Y. Saboohi, Mathematical modeling of the reaction in an iron
ore pellet using a mixture of hydrogen, water vapor, carbon monoxide and carbon dioxide: an

IP
isothermal study, Advanced Powder Technology. 17 (2006) 277-295.
[4] A. Pineau, N. Kanari, I. Gaballah, Kinetics of reduction of iron oxides by H2 Part II. Low

R
temperature reduction of magnetite, Thermochimica Acta. 456 (2007) 75-88.

SC
[5] S. Halder, R.J. Fruehan, Reduction of Iron-Oxide-Carbon Composites: Part II. Rates of
Reduction of Composite Pellets in a Rotary Hearth Furnace Simulator, Metallurgical and
Materials Transactions B. 39 (2008) 796-808.

NU
[6] S. Halder, R.J. Fruehan, Reduction of Iron-Oxide-Carbon Composites: Part I. Estimation of the
Rate Constants, Metallurgical and Materials Transactions B. 39 (2008) 784-795.
[7] S. Halder, R.J. Fruehan, Reduction of Iron-Oxide-Carbon Composites: Part III. Shrinkage of
MA
Composite Pellets during Reduction, Metallurgical and Materials Transactions B. 39 (2008)
809-817.
[8] K. Piotrowski, K. Mondal, H. Lorethova, L. Stonawski, T. Szyma´nski, T. Wiltowski, Effect of
gas composition on the kinetics of iron oxide reduction in a hydrogen production process,
D

International Journal of Hydrogen Energy. 30 (2005) 1543-1554.

TE

[9] B. Weiss, J. Sturn, S. Voglsam, F. Winter, J. Schenk, Industrial fluidized bed direct reduction
kinetics of hematite ore fines in H2 rich gases at elevated pressure, Chemical Engineering Science.
66 (2011) 703-708.
P

[10] S.Y. Luo, C.J. Yi, Y.M. Zhou, Direct reduction of mixed biomass-Fe2O3 briquettes using
CE

biomass-generated syngas, Renewable Energy. 30 (2011) 3332-3336.

[11] P. Palacios, M. Toledo, M. Cabrera, Iron ore reduction by methane partial oxidation in a
porous media, International Journal of Hydrogen Energy. 40 (2015) 9621-9633.
AC

[12] D.B. Guo, M. Hu, C.X. Pu, B. Xiao, Z.Q. Hu, S.M. Liu, X. Wang, X.L. Zhu, Kinetics and
mechanisms of direct reduction of iron ore-biomass composite pellets with hydrogen gas,
International Journal of Hydrogen Energy. 40 (2015) 4733-4740.
[13] L.Y. Yi, Z.C. Huang, T. Jiang, L.N. Wang, T. Qi, Swelling behavior of iron ore pellet reduced
by H2-CO mixtures, Powder Technology. 269 (2015) 290-295.
[14] S. Hayashi, Y Iguchi, Factors Affecting the Sticking of Fine lron Ores during Fluidized Bed
Reduction, ISIJ International. 32 (1992) 962-971.
[15] S. Hayashi, S. Sawal, Y Iguchi, Influence of Coating Oxide and Sulfur Pressure on Sticking
during Fluidized Bed Reduction of Iron Ores, ISIJ International. 33 (1993) 1078-1087.
[16] M. Bartels, W.G. Lin, J. Nijenhuis, F. Kapteijn, J.R. Ommen, Agglomeration in fluidized beds
at high temperatures: mechanisms, detection and prevention, Progress in Energy and Combustion
Science. 34 (2008) 633–666.
[17] B. Zhang, X.Z. Gong, Z. Wang, Z.C. Guo, Relation between Sticking and Metallic Iron
Precipitation on the Surface of Fe2O3 Particles Reduced by CO in the Fluidized Bed, ISIJ
International. 51 (2011) 1403-1409.
[18] L.Y. Yi, Z.C. Huang, T. Jiang, Sticking of iron ore pellets during reduction with hydrogen and
 ACCEPTED MANUSCRIPT

carbon monoxide mixtures: Behavior and mechanism, Powder Technology. 235 (2013)
1001–1007.
[19] J.X. Li, R.F. Wei, H.M. Long, P. Wang, D.Q. Cang. Sticking behavior of iron ore-coal pellets
and its inhibition, Powder Technology. 262 (2014) 30-35.
[20] N.A. El-hussiny, M.E.H. Shalabi, A Self-Reduced Intermediate Product from Iron and Steel

T
Plants Waste Materials Using a Briquetting Process, Powder Technology. 205 (2011) 217-223.

R IP
SC
NU
MA
D
P TE
CE
AC
 ACCEPTED MANUSCRIPT

T
RIP
SC
NU
MA
Figure 1
D
TE
P
CE
AC
 ACCEPTED MANUSCRIPT

T
RIP
SC
NU
MA
Figure 2
D
TE
P
CE
AC
 ACCEPTED MANUSCRIPT

T
RIP
SC
NU
MA

Figure 3
D
TE
P
CE
AC
 ACCEPTED MANUSCRIPT

T
RIP
SC
NU
MA
Figure 4
D
TE
P
CE
AC
 ACCEPTED MANUSCRIPT

T
RIP
SC
Figure 5
NU
MA
D
TE
P
CE
AC
 ACCEPTED MANUSCRIPT

Table 1
Chemical composition of iron ore
Composition Fe total FeO SiO2 CaO Al2O3 MgO Na2O K2O TiO2 S P

Mass% 66.8 26.1 4.39 0.17 0.62 0.51 0.011 0.012 0.062 0.053 0.039

T
R IP
SC
NU
MA
D
P TE
CE
AC
 ACCEPTED MANUSCRIPT

Table 2
Selected kinetics models for reduction reaction
Kinetics models Equation

1  1   
three-dimensional phase-boundary controlled reaction (5)
 kt
13

1  1    

T
three-dimensional diffusion controlled reaction 13 2 (6)
 kt

R IP
SC
NU
MA
D
P TE
CE
AC
 ACCEPTED MANUSCRIPT

Table 3
The calculated results in CO and H2 atmosphere
Atmosphere Model T (℃) k(s-1) R MD RSMD
CO atmosphere three-dimensional phase-boundary 900 9.99E-05 0.992 -0.0570 0.0605
controlled reaction 1000 1.83E-04 0.972 -0.0937 0.1078

T
1100 3.08E-04 0.961 -0.1231 0.1457
three-dimensional diffusion controlled 900 2.23E-05 0.993 0.0474 0.0531

IP
reaction 1000 7.22E-05 0.998 0.0561 0.0765
1100 2.07E-04 0.992 0.0616 0.1038

R
H2 atmosphere three-dimensional phase-boundary 900 1.19E-04 0.993 -0.0598 0.0636

SC
controlled reaction 1000 2.17E-04 0.978 -0.0890 0.1032
1100 3.70E-04 0.980 -0.0906 0.1112
three-dimensional diffusion controlled 900 3.10E-05 0.991 0.0568 0.0639

NU
reaction 1000 9.97E-05 0.998 0.0703 0.0931
1100 2.87E-04 0.998 0.0886 0.1324
MA
D
P TE
CE
AC
 ACCEPTED MANUSCRIPT

T
IP
R
SC
NU
Graphical abstract MA
D
P TE
CE
AC
 ACCEPTED MANUSCRIPT

Highlights
 A thermogravimetric method was used to investigate the reduction experiments.
 The reaction kinetics parameters were estimated based on different model equations.
 The phase transformation of the reduction process was studied by XRD.
 The microstructure of the reduction process was studied by SEM.
 The effect of gas composition on reduction characteristics was investigated.

T
R IP
SC
NU
MA
D
P TE
CE
AC