refractory and insulator, this reactor contains a bed of iron ore pellets which will ...
presented a model which was developed to simulate transitory condition, after ...
SIMULATION OF IRON ORE REDUCTION IN A FIXED BED J. Aguilar, R. Fuentes, R. Viramontes
*
Universidad Autónoma de Nuevo León Facultad de Ingeniería Mecánica y Eléctrica Doctorado en Ingeniería de Materiales A.P. 076 "F", Monterrey, N.L. 66450 México *
HYLSA, S.A. de C.V.
Departamento de Investigación y Desarrollo A.P. 996, Monterrey, N.L. 64000, México
1. ABSTRACT A simulation of direct reduction in a fixed bed process of iron ore is presented. Simulation is done with a model which takes into account bed characteristics, including reaction kinetics, thermal
effects,
quality
and
flow
of
reducing
gas.
It
is
considered that reactor is a cylinder with wall made of layers of refractory and insulator, this reactor contains a bed of iron ore pellets which will be reduced under a flow of reducing gas in longitudinal direction (along cylinder). Finite differential
difference equations
method for
is
used
description
to of
solve
heat
simplified
transfer
gas-
pellet, gas-refractory, refractory-insulator-environment, and heat changes due to reaction inside reactor. 1
Simulator is fitted by using reducibility tests in laboratory and runs in a pilot plant. It is found that due to process nature, this process is carried out almost fully in transitory stage and that is way this simulation is necessary.
2. INTRODUCTION At
present
time,
direct
reduction
process
are
calling
researchers [1,2,3,4,5] attention in world, whom are looking for an efficient reduction process which includes advantages of direct reduction together with smelter reduction in a blast furnace, this process has been called "Smelter Reduction Process". Taking into account that there are process developed until commercialization,
and
world
trend
is
still
including
direct
reduction as part of a new process, becomes important to study this stage by using available tools. One of most common low cost tools used for design and test of new plant is simulation processes. In case of chemical processes which are batch conducted, one of the most important knowledge is time necessary to carried out such process, that is the aim of this work, to determinate time necessary to reduce an iron ore bed in contact with reducing gas and to obtain a desired reduction degree (in this case to wustite) under given operation conditions. Fixed bed scheme is more complex than moving bed, because of the
first
presented
one a
includes
model
which
transitory was
state.
developed
to
In
this
simulate
paper
is
transitory
condition, after this it is uncomplicated to pass to moving bed. 2
3. GENERAL DESCRIPTION OF A NEW REDUCTION PROCESS In a general way, new reduction process consists of a smelter unit charged with pre-reduced iron ore that is reduced to iron by combustion of carbon and oxygen. Combustion products have enough reducing potential to pre-reduce iron ore to wustite, that is way the gases are passed through a reactor with a bed of iron ore pellets which will be pre-reduced by direct reduction process and charged to smelter unit. With the goal of evaluate this process some researchers [6,7] have ran tests on pilot plant, however this model only simulates direct reduction part (pre-reduction).
4. DEVELOPMENT It is assumed in this work that reactor is a cylinder with compound wall made of refractory and insulator layers (Figure 1), iron ore passing
(pellet) through
is
inside
from
top
of to
cylinder
and
reducing
bottom.
Gas
is
gas
flowing
is in
unidirectional way following z axe in a plate front (turbulent flow). Used equations are explained below: 4.1
Heat transfer This concept corresponds to heat transfer between gas and
pellet,
and
gas
and
refractory-insulator
system
in
wall
and
environment. This section takes into account physic properties of gas used by process.
3
Equations which handle heat transfer are: a)
Heat transfer from gas;
∂ T g H gp Ap ( T g - T p ) + H g Ar ( T g - T r ) = ∂z Cg M g
b)
1
Heat transfer to pellets;
∂ T p H gp Ap ( T g - T p ) = ∂t Cp M p
c)
Heat
transfer
between
insulator
and
2
refractory
layers
in
wall;
∂T r = pra4 T f + pra5 ( T r - T a ) + pra6 ( T a - T m ) ∂t
3
∂T a = pra1 T f + pra2 ( T r - T m ) + pra3 ( T a - T m ) ∂t
4
with heat transfer coefficients for a pellet bed;
3
C1 =
ρ ( a) ρp
ρ a D (1 - a ) ρp 0.3
5
1.4
4
1 g G g 0.7 W = ( ) HH g =gpC 1 α g 1 µag + Hg 5 Kp
7 6
and an auxiliary function for transitory stage in wall.
T f = Hg εb
4.2
(T g -Tr )
8
Kb
Kinetics of iron ore reduction This
inherent
aspect to
composition
is
process are
very due
important to
strongly
thermal
related,
because profile in
other
it
is
and
an
element
chemical
words,
gas
reduction
kinetics depends on temperature and gas composition, and these depend again on reduction kinetics. Reduction
kinetics
employed
in
this
work
includes
three
oxidation stages of iron, in a simultaneous reduction process where first reduction product, which corresponds to first oxide, is reactive for next reaction for a homogeneous process. Equations for this section are related to two reactions, the first one is reduction by hydrogen, and second one is reduction by carbon monoxide. Reduction from hematite to magnetite (Fe2O3 to Fe3O4) is described as;
5
∂ Rm = Rm + Rm ∂t h
c
9
hydrogen contribution is governed by;
Rm = Am ρ g H 2 (1 - Rm )
10
Qmh ) R Tp
11
h
h
and;
Am = fmh exp(h
Same procedure is used for CO (in equation (10) term H2 related with hydrogen pressure is exchanged by CO pressure). From reduction from magnetite to wustite (Fe3O4 to FeO) and from wustite to iron (FeO to Fe) expressions are very similar, but in these cases thermodynamic constants for equilibrium are taken into account, this constants are almost zero for reduction from hematite to magnetite (that is why for hematite to magnetite is independent of water pressure), besides it is necessary to take into account the reagent amount given by reaction just before, this consideration gives the following equation for magnetite to wustite:
Rw = Aw ρ g h
h
( H 2 - K w H 2 O) ( Rm - Rw ) 1+ K w h
12
h
and for wustite to metallic iron;
6
R f = Af ρ g h
h
( H 2 - K f H 2 O) ( Rw - R f ) 1+ K f h
13
h
Reduction rate is handle in same way that reaction from hematite to magnetite, and with CO contributions (same kind that hydrogen ones) there are enough equations to know global reduction degree. Thermal behavior is taken from heat of reaction of each one of involved reaction, this contribution is added to heat transfer section. Even when the objective of this work is centered on wustite production, this model takes into account reduction from iron ore to metallic iron because some times it is possible to have conditions, which permits to reach metallic iron.
4.3 Kinetics of gas conversion In
this
section
equilibrium
degree
inside
of
bed
is
considered, with this information it is possible to work with thermodynamic equilibrium letting gases to complete reactions, or as an opposite way, to inhibit gas reactions away of thermodynamic equilibrium. With these capacities the model can reproduce any experimental conditions and know how far from equilibrium the gasgas reactions are taking place.
4.4
Condensation and evaporation of water This phenomenon must be considered especially in cases where
reducing gas
composition
is
high
in
hydrogen
and
water
is
a
7
reaction product. This is also very important because simulation and actually the process starts from room temperature and water vapor
is
condensing,
when
the
iron
ore
bed
reaches
water
evaporation temperature water boils and modify the temperature profile
strongly.
This
phenomena
is
handling
by
taking
into
account pressure inside reactor and pellet temperature, thus water change its state from liquid to vapor and backward, water amount in
this
changing
obviously it
can
process not
be
depends grater
on than
evaporation energy
kinetics
provided
by
and heat
transfer gas-pellet. This part is also included in heat transfer section with heat transfer terms in equations.
5. EQUATIONS RESOLUTION PRESENTED AS SIMULATION MODEL Flow diagram in figure 2 shows the sections above described linked with the aim to have a fully integrate simulator. Start point
is
to
have
general
initialization
values,
such
as
ore
properties, process parameters (temperature, composition and flow of reducing gas), reactor characteristics (dimensions and physic properties of refractory and isolator) and data related to iron ore bed to be reduced. Due to fixed bed is a batch process, thermal transitory state is taking place while reduction (required time for reduction is less than time for reach steady state) and this condition makes necessary
to
evaluate
and
take
into
consideration
thermal
parameters of refractory-insulator-environment [8]. 8
In
accordance
with
equations
showed
before,
finite
differences method uses a loop in time which contains a loop along z axe.
Adjust of right parameters in model was done by runs in
laboratory
and
pilot
plant.
Simulation
model
reproduces
experimental runs with different conditions, thus it is already validated.
6. EXPERIMENTAL 6.1 Experimental design for laboratory tests Experimentation was conducted on a fixed mass of 500 gr. of o
o
iron ore pellets with three different temperatures (750 C, 850 C o
and 950 C) and three reducing gasses of different compositions where employed, gas flow was 55 lts/min. Experimental design for laboratory test is listed in tables I through IV: Table I Design Matrix for reduction from hematite to magnetite Tempera o ture C
H2/H2O (1)
H2/H2O (2)
H2/H2O (3)
CO/CO2 (1)
CO/CO2 (2)
CO/CO2 (3)
750
0.1
0.37
0.64
0.1
0.28
0.47
850
0.1
0.22
0.34
0.1
0.22
0.36
950
0.1
0.15
0.20
0.1
0.19
0.29
Table II Design Matrix for reduction from magnetite to wustite
9
Tempera o ture C
H2/H2O (1)
H2/H2O (2)
H2/H2O (3)
CO/CO2 (1)
CO/CO2 (2)
CO/CO2 (3)
750
0.70
0.40
0.22
0.51
0.39
0.31
850
1.41
1.09
0.90
1.05
1.16
1.28
950
2.18
1.84
1.60
1.63
1.97
2.28
Table III Design Matrix for reduction from wustite to iron Tempera o ture C
H2/H2O (1)
H2/H2O (2)
H2/H2O (3)
CO/CO2 (1)
CO/CO2 (2)
CO/CO2 (3)
750
2.39
2.02
1.76
1.79
2.16
2.50
850
4.13
3.56
3.18
3.05
3.78
4.54
950
6.08
5.29
4.77
4.47
5.64
6.84
H2/H2O FeO to Fe
CO/CO2 FeO to Fe
Table IV Complementary Matrix design Tempera o ture C
H2/H2O Fe2O3 to Fe3O4
650
0.69
800
0.23
0.21
1.21
1.11
∞
∞
950
0.10
0.14
0.89
1.29
∞
∞
CO/CO2 Fe2O3 to Fe3O4
H2/H2O Fe3O4 to FeO
CO/CO2 Fe3O4 to FeO
∞
2.00
6.2 Ore Tested ore comes from the Alzada mine, located at the Mexican State of Colima. The mean characteristics of this ore are given in Table V.
10
Table V Mean ore characteristics PHYSICS Granulation
+ 3/8" - 5/8" (0.95-1.58 cm)
Average diameter
1.27 cm
Density
4.222 gr/cm
Apparent density
2.200 gr/cm
3
3
CHEMICAL Total Fe
66.5 %
+2
Fe
0.65 %
Gangue
5.3 %
CaO
37 %
MgO
11 %
SiO2
38 %
Al2O3
14 %
6.3 Experimental devices and procedures Experimental
tests
reducibility
tests.
reactors
conduct
to
The
were
conducted
equipment
iron
ore
in
includes
reduction
an
and
a
rector
18
Kw
control
for
furnace, it,
and
instrumentation, as it is shown in Figure 3. This laboratory is fully prepared to make comparative reducibility ore tests. It is possible to change the temperature, gas flow, pressure and charge to which the pellets are exposed to. A bed of 500 gr. of dry ore pellet were placed inside a 11
reactor. This reactor consists in two concentric tubes in such a way that the inlet gas pass between the internal wall of the external tube and the external wall of the internal tube, and leaves it through the ore bed placed inside of the internal tube (Figure 4). The reactor has a thermocouple in contact with the top of the ore bed and another on bottom for temperature controlling. It is important to notice that this experimental device gives information of an iron ore bed reduced in batch process, and simulation of bed conditions is an aim of this work. Searching for constants is the first application of simulation model. In order to
simplify
search
for
constants,
each
reducing
stage
was
conducted independently (Tables I through IV) so there is just one set of constants to seek in each stage. A 10 lts/min flow of high purity nitrogen is passed through the sample until the testing temperature is reached and stabile, after this is done, the appropriate gas is injected and the weight loss is registered. The flow control is made by calibrated flowmeters and the gas composition is checked by chromatography analysis. The water vapor amount in gas is indirectly measured. During the reduction the sample looses weight until all the oxygen has been removed. In this kind of test, the sample is heated by radiation and conduction. In order to simulate a real bed, in a second part alumina pellets where placed on top and bottom of the ore bed, while the reactor walls where covered with an insulating wool. All these avoids the instability of the gas 12
flow entering into the bed, some results are presented in figures 5 through 9.
6.4 Experimental procedures for pilot plant tests Pilot plant consist in a vertical steel cylindrical reactor with wall covered of refractory and insulator layers, inlet gas which passes through pellet bed, is at top and wasted gas leaves reactor at bottom. There are several thermocouples and takes for gas analysis along reactor. Reactor is fully instrumented with central data acquisition and control for temperatures, pressures, gas flows, gas compositions. General characteristics of this reactor are shown in Table VI, this information is provided to simulation model. Cages of ore sated at same positions than thermocouples and test points for gas analysis, these cages are recovered after tests for a chemical analysis in order to have thermal and chemical data. Preparation for test consists of charging reactor. Depending on the kind of test it is possible to heat the ore bed with hot nitrogen,
air
or
a
selected
gas
until
it
reach
desired
temperature. If a reducing gas is used, then a reduction will start too. Timing for tests in this work starts when gas is getting in reactor, thus heating and reduction starts at same time. Selected gases are shown in Table VII. Most used gas is shown in column three. Normal
criteria
to
define
the
end
of
process
is
taking
temperature at bottom of bed, looking evolution of gas composition 13
or flows between inlet gas and outlet gas.
Table VI General Data Item
Value
Density of refractory
2300 Kg/m
Refractory's heat capacity
1178 J/Kg C
Density of insulator
800 Kg/m
Insulator's heat capacity
1154 J/Kg C
Apparent density of bed
1900 Kg/m
Density of a single pellet
4222.2 Kg/m
Pellet's heat capacity
957.6 J/Kg C
Reactor's diameter
0.914 m
3
o
3
o
3
3
o
-3
Radius of a single pellet
6.35 x 10
m
Thickness of refractory
0.0762 m
Thickness of insulator
0.1016 m
Pellets' thermal conductivity
0.5 W / m C
Refractory's thermal conductivity
1.441 W / m C
Insulator's thermal conductivity
0.255 W / m C
Refractory and insulator masses
1.749 & 0.956 m tons
Pellet mass
3.4 m tons
Height of bed
3.20 m
o
o
o
Table VII Chemical analysis of used gases for reduction and operation temperatures Temp.
o
o
o
o
o
950 C
950 C
900 C
800 C
800 C
23.12
14.8
24.14
27.14
48.80
vol % H2
14
CO
10.61
15.8
10.52
10.19
16.06
CO2
04.35
09.60
04.53
05.15
08.09
H2O
15.29
12.80
15.23
14.71
26.56
N2
46.60
46.00
45.57
42.77
00.27
CH4
< 0.01
01.00
< 0.01
0.034
0.123
10. RESULTS AND DISCUSSION
10.1 Simulator adjustment Simulator is based on basic principles, which governs the process behavior by using a set of functions and equations with constants
experimentally
determined.
First
part
of
simulator
adjustment is to take constant values from those already published [9,10,11,12,13,14,15,16] for heat transfer in beds, on the other hand,
reducibility
tests
at
laboratory
are
used
to
determine
kinetic constants for iron ore reduction following those above proposed. With
this
set
of
parameters
simulator
is
tested
by
comparisons between model results with actual pilot plant data. Constant values were adjusted until model reproduced experimental results. Constants has a physical sense which not was changed along adjusting, constants still having their sense and magnitude order that represent. These constants are including effects that are
not
numbers,
described
separately,
which
related
convective
are
transfer,
and
such
with
Knudsen
bed
as and
Sherwood flow
diffusivity
and
Schmidt
conditions
for
coefficient
for 15
diffusion
in
a
pellet.
Model
presented
in
this
work
is
a
simplification, which reproduces experimental data and gives a very straight criteria for determining the end of process. One of the experimental runs is showed on Figure 10, which presents thermal profile of reactor at the different thermocouples position along z axe, and its evolution with time, it is possible to
see
that
temperature
at
the
beginning
of
test
is
almost
constant around boiling temperature water, thus it is shown that it is very important to consider condensation - boiling phenomena inside model. Heat transfer coefficients inside simulation model are moving in order to achieve the same temperature profiles obtained at pilot plant, and it is necessary to take into account that could be interactions between phenomena which are not considered in mathematical
model,
but
they
are,
as
it
was
pointed
before,
included inside of constants. Boundary phenomena such as diffusion of gas in the layer of reaction products and interaction of the phase-boundary reaction with diffusion in the reacting layers are in adjusted constants because there are many effects that can not be evaluated in a single way (combined effects of the phaseboundary reaction, gas flow, diffusion in reacting layers, and diffusion in the end products reaction). Kinetic constants were evaluated from reducibility tests and cages located inside reactor in know places (same positions than thermocouples). Kinetic reductions are strongly important because it
has
influence
on
temperature
profiles
and
gas
conversion 16
kinetics. Kinetics
of
reduction
gas
conversion
was
calculated
from
analysis of samples of gas obtained along reactor at the same positions
than
thermocouples,
this
kinetic
was
adjusted
until
model gives the same results than chemical analysis. All of the profiles (bed temperature, wall temperature, iron ore reduction degree and chemical gas composition) were reproduced. When
constants
satisfy
all
of
these
conditions
then
the
simulation model is complete and can be used, curves in Figure 11 were obtained using the simulator and the results are in good agreement with actual data. With this model, next step is to calculate reduction time, and with this information it is possible to
calculate
specific
consumption
of
gas
for
certain
given
conditions, which helps to give an advice on operation conditions.
10.2 Simulator validation Due
to
the
objective
of
this
work
is
oriented
to
make
predictions about reduction time, at this point, with specific operation
conditions,
different
than
those
used
for
model
adjustment, this simulator gives results which had been compared with
tests
agreement. function
carried Figure
of
out
12
reducing
in
shows gas
pilot
plant
a
curve
flow
at
for one
and
exhibits
reduction given
a
time
temperature
good as
a
and
composition, this curve was obtained by using the model, and experimental data from runs are plotted. Notice that when gas flow is not enough to compensate heat loss the time to finish reduction 17
is too long, and when gas flow is very high there is a kinetic frontier imposed by process. With
this
information
and
including
the
reducing
gas
generation it is possible to get specific consumption. Some adjusted parameters and functions are presented in Table VIII
through
XI,
gas
properties,
thermal
and
thermodynamic
constants are already published in several books and papers. Previous work [17] with this kind of atmosphere shows that reduction with carbon monoxide is about one fifth of hydrogen, thus
in
this
case
water
shift
reaction
outside
pellet
is
considered.
Table VIII Adjusted parameters and functions (Kinetics) Parameters and functions
Am = 45 exp(h
6000 Tp
)
Remarks Kinetic constant for reduction from hematite to magnetite with hydrogen
18
Aw = 0.36 exp(-
2600
h
A f = 0.18 exp(-
Tp
2600
h
Tp
)
Kinetic constant for reduction from magnetite to wustite with hydrogen
)
Kinetic constant for reduction from wustite to metallic iron with hydrogen
h
Kinetic constant for reduction from hematite to magnetite with carbon monoxide
Aw Aw = 5
h
Kinetic constant for reduction from magnetite to wustite with carbon monoxide
Af Af = 5
h
Kinetic constant for reduction from wustite to metallic iron with carbon monoxide
Am Am = 5 c
c
c
Table IX Adjusted parameters and functions (Kinetics) Parameters and functions
Remarks Reduction rate for hematite to magnetite with hydrogen
Rm = Am ρ g H 2 (1 - Rm ) h
h
( Rm - R w ) Rm = Aw ρ g ( H 2 - K w H 2 O) 1+ K w h
h
h
Reduction rate for magnetite to wustite with hydrogen
h
19
( Rw - R f ) R f = A f ρ g ( H 2 - K f H 2 O) 1+ K f h
h
h
Reduction rate for wustite to metallic iron with hydrogen
h
Reduction rate for hematite to magnetite with carbon monoxide
Rm = Am ρ g CO (1 - Rm ) c
c
( Rm - Rw ) Rw = Aw ρ g (CO - K w CO 2 ) 1+ K w
Reduction rate for magnetite to wustite with carbon monoxide
( Rw - R f ) R f = A f ρ g (CO - K f CO 2 ) 1+ K f
Reduction rate for wustite to metallic iron with carbon monoxide
c
c
c
c
c
c
c
c
Table X Adjusted parameters and functions (Gas-gas reactions) Parameters and functions
λ = 2 exp(-
9000 Tp
)
K shift CO H 2 COequilibrium = H2O
Remarks Apparent kinetic constant for water shift reaction CO+H2O=CO2+H2
CO calculated from equilibrium of water shift reaction
20
CO - COequilibrium CO non-equilibriumdegree = CO + COequilibrium
β = 0.04 exp(-
12000 Tp
)
CO non equilibrium degree in accordance with water shift reaction
Apparent kinetic constant for methane reforming reaction CH4+H2O=3H2+CO
All of this reactions have a thermal contribution on system, this
contribution
is
integrated
to
mass
and
heat
balance
equations.
Table XI Adjusted parameters and functions (Gas-gas reactions) Parameters and functions
CH 4
CH 4
equilibrium
2 CO H 32 P = H 2 O K reforming
CH 4 - CH 4 = CH 4 + CH 4
equilibrium
non-equilibriumdegree
Remarks CH4 calculated from equilibrium of methane reforming reaction
CH4 non equilibrium degree in accordance with methane reforming reaction
equilibrium
21
CO coming from water shift reaction
GIP = C 1 ρ g λ CO non-equilibriumdegree 2
BETAP = C 1 ρ g β CH 4
CH4 coming from methane reforming reaction
2
non-equilibriumdegree
11. CONCLUSION
When reproduction of experimental data is achieved, including thermo-chemical profiles inside reactor and time to reduction, it is possible to say that model is already validated and its results are in this scope reliable. It is clear that simulator represents, under different conditions, the general behavior of reactor. With a model of simulation of this kind there is another potential tool for reactors design, and a very important advance in statistic experimental designs, besides, it is possible to make a
sensibility
analysis
of
different
process
variables
and
operational conditions. Even when by adjusting model it is not possible to give values
for
resistance),
each
phenomena
adjusted
in
pellet
parameters
gave
and good
bed
(i.e.
results
diffusion about
bed
behavior. Further studies will permit to determine the limitations of
this
simulation
model
and
if
it
is
necessary
to
made
modifications.
22
Finite
difference
method
gives
the
opportunity
to
pass
directly to moving bed by changing elements conditions and places. This model gives a criteria to determinate the end of a batch process,
also
gives
reducing
gas,
and
information the
chance
about to
specific
modify
and
consumption test
of
different
atmospheres at different temperatures.
11. NOMENCLATURE ais the pellet radius. Amhis the kinetic term for reduction from hematite to magnetite with hydrogen. Amcis the kinetic term for reduction from hematite to magnetite with carbon monoxide. Afh and Awh are kinetic terms for reduction from wustite to iron and from magnetite to wustite respectively with hydrogen. Afc and Awc are kinetic terms for reduction from wustite to iron and from magnetite to wustite respectively with carbon monoxide. Ap and Ar are respectively pellet area and refractory area. Cg and Cp are respectively heat capacities of gas and pellet. COis carbon monoxide partial pressure. Dis the reactor diameter. Ggis the gas flow. 23
H2is hydrogen partial pressure. H2Ois partial pressure of water vapor. Hg and Hgp are coefficients for heat transfer in the iron ore bed and gaspellet. Kb and Kp are coefficients for heat conduction in refractory and pellet. Kwh, Kfh, Kwc and Kfc are
thermodynamic
constants
for
reduction
from
magnetite
to
wustite and wustite to metallic iron with hydrogen and with carbon monoxide respectively. Kshiftequilibrium constant for water shift reaction. Kmethane equilibrium constant for methane reforming reaction. Kfh and Kwh are thermodynamic equilibrium constant for reduction from wustite to iron and from magnetite to wustite respectively, with hydrogen.
Mg and Mp are the masses of gas and pellet. prai are
coefficients
for
heat
transfer
coefficient
of
refractory-
insulator-environment system. Qmhrepresents activation energy for reduction from hematite to magnetite with hydrogen. 24
Rm, Rmc and Rmh are reduction degree for hematite to magnetite, reduction degree due to carbon monoxide and hydrogen. Rf and Rw are reduction degrees of reaction from wustite to iron and from magnetite to wustite. Ta, Tg, Tf, Tm, Tp and Tr are
respectively
temperatures
of
insulator,
gas,
system
wall,
environment, pellet and refractory. tis the time coordinate Wgis the molecular weight of gas zis the coordinate along reactor cylinder αgis the heat conduction coefficient of gas εbis the refractory thickness ρais the apparent density of iron ore bed ρpis the pellet density μgis the gas viscosity. ρgis reducing gas density.
ACKNOWLEDMENTS J. A. gives the thanks to the CONACYT (National Science and Technology Council) for its support along this work.
REFERENCES 25
1-Lascano (A.), Villaseñor (A.), Alcántara (M.).- Alternativas de integración
de
la
reducción
directa
a
un
convertidor
inyectando oxígeno y carbón por el fondo dando como resultado un nuevo proceso IMIS de producción de acero. V Seminario Reducción
Directa,
Saltillo,
Coah.
México,
14
al
16
de
octubre, 1986.: ILAFA, 1986, p. M1-M7 2-Flickenschild (J.), Papst (G.).- El proceso COREX de fusión reductora. V Seminario Reducción Directa, Saltillo, Coah. México, 14 al 16 de octubre, 1986.: ILAFA, 1986, p. L1-L21 3-González (F.).- Perspectivas futuras del uso del hierro esponja en la producción de acero. V Seminario Reducción Directa, Saltillo, Coah. México, 14 al 16 de octubre, 1986.: ILAFA, 1986, p. B1-B2 4-Quintero
(R.G.).-
Up
Date
on
HyL.
ISS-AIME
Ironmaking
Proceedings, 1978, v. 37, p. 137-148 5-Peña (J.M.), Viramontes (R.).- The New HyL Technology. ISS-AIME Ironmaking Proceedings, 1980, v.39 6-Lascano (A.).- Progress made in the IMIS process for direct steelmaking. Steel Times International, January, 1990, p.3638 7-Nijhawan (B.R.).- Desarrollos en la reducción directa y nuevos procesos para la fabricación de acero, una evaluación. V Seminario Reducción Directa, Saltillo Coah. México, 14 al 16 de octubre, 1986.: ILAFA, 1986, p. C1-C10 8-Santillán
(M.).-
Modelo
matemático
lineal
aproximado
en
la
transferencia de calor (Caso unidimensional). Science Master 26
Degree Tesis, FIME-UANL, México, June 1992 9-Bogdandy (L.), Engell (H.J.).- The Reduction of Iron Ores.-New York: Springer-Verlag, 1971, p.576 10-Stull
(D.R.),
Prophet
(H.).-
JANAF
Thermochemical
Tables,
Second Edition: National Standard Reference Data System 11-Moelwyn (H.).- The kinetics of reactions in solution, Second Edition.- Clarendon: Press Oxford, 1947 (W.K.),
12-Lu
Bitsianes
reduction
of
(G.).-
hematite.
Chemical
Canadian
kinetics
of
Metallurgical
gaseous
Quarterly,
v.7,nº1 13-Seth (B.B.L), Ross (H.U.).- Application of a generalized rate equation to the gaseous reduction of iron oxide. Canadian Metallurgical Quarterly, v. 5, nº 4, p. 315-328 14-The
making
shaping
and
treating
of
steel
10th
edition:
Association Of Iron And Steel Engineers, 1988 15-Fuentes (R.), Aguilar (J.) Núñez (C.), et al..- Heat transfer and reduction to wustite of a bed of hematite pellets, Report DIM/10/02-040 HYLSA, June, 1989. 16-Fuentes (R.), Aguilar (J.) Farías (L.).- Mathematical modelling of
the
behavior
of
different
gases
for
reduction
from
hematite to wustite in the HYLSA pilot fixed bed plant, Report DIM/01/06-073 HYLSA, November, 1989. 17-Fuentes
(R.),
Aguilar
(J.).-
Reduction
from
hematite
to
wustite, Report DIM/01/04-22, November, 1988.
27
28
Figure
1.Scheme of a reactor of fixed bed reduction represents thermocouple positions along reactor)
(TR-n
29
Figure
2.Flow diagram process
for
the
simulation
of
direct
reduction
30
Figure 3. Reducibility Laboratory
31
Figure 4. Reducibility reactor scheme
32
Figure 5. Experimental data reduction from hematite to magnetite o at 650 C with H2/H2O = 41/59
33
Figure 6. Experimental data reduction from magnetite to wustite at o 750 C with H2/H2O = 58/42
34
Figure 7. Experimental data reduction from wustite to metallic o iron at 850 C with H2/H2O = 84/16
35
o
Figure 8. Reducibility test, 800 C, 55 liters/minute, (55% H2, 21% CO, 14% CO2, 10% N2)
36
o
Figure 9. Reducibility test, 950 C, 55 liters/minute (55% H2, 21% CO, 14% CO2, 10% N2)
37
Figure
10.Reduction test of pellet charge of 4400 Kg.
from
hematite
to
wustite,
38
Figure 11.Simulation of run showed in Figure 10
39
Figure
12.Reduction time against flow of composition than figure 11)
reducing
gas
(same
40
