Proposal Menara Masjid
Proposal Menara Masjid
Proposal Menara Masjid
www.elsevier.nl/locate/ces
Received 28 April 2000; received in revised form 13 September 2000; accepted 2 December 2000
Abstract
For combustion with CO2 capture, chemical-looping combustion has the advantage that no energy is lost for the separation of
CO2 . In chemical-looping combustion oxygen is transferred from the combustion air to the gaseous fuel by means of an oxygen
carrier. The fuel and the combustion air are never mixed, and the gases from the oxidation of the fuel, CO2 and H2 O, leave the
system as a separate stream. The H2 O can easily be removed by condensation and pure CO2 is obtained without any loss of
energy for separation. This makes chemical-looping combustion a most interesting alternative to other CO2 separation schemes,
which have the drawback of a large energy consumption. A design of a boiler with chemical-looping combustion is proposed. The
system involves two interconnected uidized beds, a high-velocity riser and a low-velocity bed. Metal oxide particles are used
as oxygen carrier. The reactivities needed for oxygen carriers to be suitable for such a process are estimated and compared to
available experimental data for particles of Fe2 O3 and NiO. The data available on oxygen carriers, although limited, indicate that
the process outlined should be feasible. ? 2001 Elsevier Science Ltd. All rights reserved.
0009-2509/01/$ - see front matter ? 2001 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 0 9 - 2 5 0 9 ( 0 1 ) 0 0 0 0 7 - 0
3102 A. Lyngfelt et al. / Chemical Engineering Science 56 (2001) 3101–3113
the liquid will rise slowly and ultimately reach the level needed, however, to compress CO2 into a liquid suitable
where gas bubbles are formed. However, if proper mea- for sequestration.
sures are taken to allow the CO2 liquid to mix with the The purpose of the present paper is to assess the poten-
seawater, the resulting mixture obtains a higher density tial of chemical-looping combustion — is it technically
than the seawater and instead sinks. The storage capacity and economically realistic? In order to do this, a tentative
is enormous, but local e8ects on the environment need to design of the process is made. Aspects of importance for
be investigated. judging the realism of the process are discussed.
Deep sea bottom storage: At high pressures, i.e.
3000 m below the sea level, the density of pure CO2 is
higher than that of seawater. Thus, CO2 released at these 2. Chemical-looping combustion
depths will form “lakes” on the ocean oor. The poten-
tial storage capacity is vast considering the fact that the Chemical-looping combustion has been discussed ear-
average depth of the sea is 3700 m and that more than lier in the literature as an alternative to normal combustion
50% of the earth’s surface is found at a depth of more (Richter & Knoche, 1983; Ishida & Jin, 1994; Anheden,
than 3000 m. However, local e8ects on the sea oor NPasholm, & Svedberg, 1995). The system is composed
environment are obviously inevitable. of two reactors, an air and a fuel reactor, as shown in
The estimated cost of CO2 disposal, e.g. 4 –8 US $= Fig. 1. The fuel needs to be in a gaseous form and is intro-
ton C (Riemer, 1998), is small compared to the costs duced to the fuel reactor, which contains a metal oxide,
for separation of CO2 , which is typically in the range MeO. The fuel and the metal oxide react according to
100 –200 US $=ton C (Freund, 1998). Consequently,
(2n + m)MeO + Cn H2m → (2n + m)Me
the major problem seems to be to extract the CO2
from the fuel conversion process. The available or + mH2 O + nCO2 : (1)
proposed technologies all have the disadvantage of
The exit gas stream from the fuel reactor contains CO2
consuming large amounts of energy. The major loss
and H2 O, and almost pure CO2 is obtained when H2 O is
in eLciency arises from the energy needed for sep-
condensed. The reduced metal oxide, Me, is transferred
aration of CO2 and for compression to liquid form.
to the air reactor where the metal is oxidized according to
Typically, the energy needed for compression is about
one-fourth of the total energy needed for separa- Me + 21 O2 → MeO: (2)
tion and compression. For a coal-@red power plant,
roughly one-@fth of the electricity produced will be A full conversion from MeO to Me and back to MeO,
lost for CO2 separation and compression (Lyngfelt & as indicated by reactions (1) and (2), is not necessarily
Leckner, 1999). This decrease in eLciency alone in- obtained in a real system. The air oxidizing the metal
creases the cost for electricity production with one-fourth, produces a ue gas containing only N2 and some un-
and in addition there are costs related to separation and used O2 . Depending upon the metal oxide used, reaction
handling of CO2 . If natural gas is used as a fuel, the (1) is often endothermic, while reaction (2) exothermic.
relative loss in eLciency is somewhat smaller, but still The total amount of heat evolved from reactions (1) and
substantial. (2) is the same as for normal combustion, where the
Today, power production contributes with one-third of
the CO2 released from fossil fuel combustion world-wide
(Herzog et al., 2000). Although the cost for separation
of CO2 is substantial, power plants using fossil fuel
and CO2 capture may well be the least costly alterna-
tive for CO2 -free power production (Keith & Parson,
2000).
Chemical-looping combustion (CLC) o8ers a solution
where no energy is needed for the separation. The pro-
cess uses a solid oxygen carrier to transfer the oxygen
from the air to the fuel. The oxygen carrier is recycled
between a fuel reactor, where it is reduced by the fuel,
and an air reactor, where it is oxidized by the air. Thus,
the air is never mixed with the fuel, and the CO2 does
not become diluted by the nitrogen of the ue gas. The
outgoing gas from the reduction step will contain water
vapour and CO2 . The water vapour can easily be sep-
arated by condensation, and the CO2 is delivered with- Fig. 1. Chemical-looping combustion. MeO=Me denote recirculated
out an energy penalty for the separation. Energy is still oxygen carrier solid material.
A. Lyngfelt et al. / Chemical Engineering Science 56 (2001) 3101–3113 3103
oxygen is in direct contact with the fuel. The advantage of from the oxygen carrier to the fuel. The volumetric gas
chemical-looping combustion compared to normal com- ow in the air reactor is approximately 10 times larger
bustion is that CO2 is not diluted with N2 but obtained than that of the gaseous fuel, and to keep a reasonable
in a relatively pure form without any energy needed for size of the reactors a high velocity is chosen in the air
separation. reactor.
Originally, the process was proposed as a method to The gas velocity in the riser provides the driving
enhance the thermal eLciency of fuel combustion. The force for the circulation of particles between the two
idea was to supply low-temperature heat to the endother- beds. Thus, the particles carried away from the riser
mic reaction in the fuel reactor, thereby increasing the are recovered by a cyclone and led to the fuel reactor.
amount of heat produced in the high-temperature air From the fuel reactor the particles are returned to the
reactor. This possibility is not considered in the present air reactor by means of gravity; the fuel reactor is lo-
application, since it would make the process more com- cated at a suLciently high level. After condensation of
plex and also involves extra demands on the properties the water, the remaining gas, containing mostly CO2 , is
of the oxygen carrier. compressed and cooled in stages to yield liquid CO2 .
Remaining non-condensable gas from this stream, such
as unreacted methane, is recycled to the fuel reactor. In
3. Reactor design order to avoid accumulation of non-combustible gases,
such as N2 , in the recycling loop, a minor part of this
The reactors in Fig. 1 could be designed in a variety ow is bled to the air reactor. Small amounts of water left
of ways, but two interconnected uidized beds have an in the liquid CO2 have to be removed by a regenerable
advantage over alternative designs, because the process solvent to make the CO2 ow less corrosive.
requires a good contact between gas and solids as well
as a ow of solid material between the two reactors.
The system proposed is a circulating system composed 4. Oxygen carriers
of two connected uidized beds, a high-velocity riser
and a low-velocity bubbling uidized bed (Fig. 2). The The metal oxide, used as an oxygen carrier in
bed material circulating between the two uidized beds is chemical-looping combustion, must have suLcient rates
the oxygen carrier in the form of metal oxide particles. In of reduction and oxidation, at the same time as it pos-
the air reactor, or the riser, oxygen is transferred from the sesses enough strength to limit particle breakage and
combustion air to the oxygen carrier. In the low-velocity attrition. It is also an advantage if the metal oxide is
uidized bed, or the fuel reactor, oxygen is transferred cheap and environmentally sound. A number of metals
and their corresponding oxides have been mentioned in
the literature as possible candidates: Fe, Ni, Co, Cu, Mn
and Cd. At Tokyo Institute of Technology, Ishida and
co-workers have investigated the rate of oxidation and
reduction of Ni, Fe and Co (Nakano, Iwamoto, Maeda,
Ishida, & Akehata, 1986; Ishida & Jin, 1994, 1996;
Ishida, Jin, & Okamoto, 1996, 1998; Ishida, Yamamoto,
& Saito, 1999; Jin, Okamoto, & Ishida, 1998, 1999) in a
thermogravimetric analyzer (TGA) using either H2 , CO
or CH4 as fuel and air as the oxidizing gas.
In order to increase the reactivity and durability of
the oxides, the particles have often been doped with
Al2 O3 , yttria-stabilized zirconium (YSZ), TiO2 or MgO.
The particle size of the oxygen carriers has been rather
large in most of these investigations, around 2 mm in
diameter. The rates of reaction vary widely depending
upon the type of metal oxide, particle size, reduction gas
and temperature. Generally, Ni and Co and their oxides
show higher oxidation and reduction rates than Fe and a
greater durability after repeated oxidation and reduction
cycles.
Many of the reduction experiments were made at a
◦
rather low temperature, less than 900 C. The reason for
Fig. 2. Layout of chemical-looping combustion process, with two this is the desire to use the fuel reactor as a heat sink to
interconnected uidized beds. enhance the process eLciency, as mentioned above. Since
3104 A. Lyngfelt et al. / Chemical Engineering Science 56 (2001) 3101–3113
Table 1
Literature data on oxygen carriers in chemical looping combustion
◦ ◦
Carrier=support Carrier=support Red. gas Tred: ( C) Tox: ( C) Dp (mm)
(continued)
Nakano et al. (1986) Fe2 O3 , Fe2 O3 = Ni H2 700 –900 800 –1000 0.007
Fe2 O3 = Al2 O3 H2 = H 2 O
Ishida and Jin (1994)a NiO=YSZ, Fe2 O3 = YSZ H2 600,800,1000 700,900,1100 1–3
NiO Fe2 O3 = Al2 O3
Ishida et al. (1996) NiO=YSZ H2 600,800,1000 600,800,1000 1.8
Ishida and Jin (1996)b NiO NiO=YSZ H2 600 1000,1200 2
Jin et al. (1998) NiO=YSZ, Co3 O4 = YSZ, H2 ; CH4 600 1000 1.8
Fe2 O3 = YSZ CoO-NiO=YSZ
Ishida et al. (1998)c NiO=YSZ Fe2 O3 = YSZ H 2 = N2 550,600,700
NiO= Al2 O3 Fe2 O3 = Al2 O3 CO= N2 800,900
NiO= TiO2 Fe2 O3 = TiO2 CO= CO2
CO= N2 = CO2
CO= N2 = H2 O
Jin et al. (1999)d NiO= Al2 O3 , CoO= MgO H2 600 1000 2.1,1.8
NiO= TiO2 , Fe2 O3 = Al2 O3 , H2 O= CH4 700 1000
NiO=MgO, Fe2 O3 = TiO2 ,
CoO= Al2 O3 , Fe2 O3 = MgO
CoO= TiO2 ,
Ishida et al. (1999)e NiO= Al2 O3 H2 (TGA) 900 900 0.07
H2 = Ar (CR )
Hatanaka et al. (1997)f NiO CH4 400 –700 0.07
a E8ect of H2 O on oxidation, e8ect of particle size.
b ◦
No NOx formation at 1200 C.
c Study of carbon deposition.
d E8ect of pressure.
e Data from continuous CLC reactor.
f Gas measurement with gas chromatography.
Table 2
Selected literature data on conversion rates
◦
Oxidation Reduction Conversion Average Agent T ( C) dp Reference
range rate (mm)
this possibility is not regarded in the present application, cannot be immediately applied to conditions where a high
there is no reason to consider a lower temperature in the conversion of the reactant gas is necessary.
fuel reactor than in the air reactor. Hatanaka, Matsuda, and Hatano (1997) performed
Most of the studies were made with the Ni–NiO sys- cyclic reactivity investigations of NiO in a @xed-bed
tem, but some data are available for Fe2 O3 and CoO. reactor, using methane and air, where the conversion of
The reduction studies were normally made with H2 as the gases was measured but not the conversion rate of
fuel, but isolated data are available for CH4 and CO. An the oxygen carrier. Recently, the conversion rate of iron
overview of the work made is given in Table 1. Exam- oxide particles was determined as function of conversion
ples of conversion rates from these studies are shown in of the gas in a @xed-bed reactor with methane as fuel
Table 2. These rate data were obtained in TGA and they (Mattisson, Lyngfelt, & Cho, 2000a,b).
are therefore probably associated with a low conversion The process of chemical-looping combustion has been
of the reactant gas. For this reason these conversion rates demonstrated in a laboratory set-up at a temperature of
A. Lyngfelt et al. / Chemical Engineering Science 56 (2001) 3101–3113 3105
◦
1200 C, where particles of NiO with a diameter of about and the reducer are given by the pressure drops, Vpox
0:1 mm were used as oxygen carrier and the fuel was H2 and Vpred :
(Ishida et al., 1999). The question of how a CLC boiler is Vpox Aox
designed remains to be answered. First, the basic design mbed; ox = ; (8a)
g
relationships will be formulated.
Vpred Ared
mbed; red = ; (8b)
g
5. Theory where g is the acceleration of gravity. The bed heights
are obtained from the density of the solid material, , and
Assuming that the fuel is completely burnt, the fuel the voidage,
consumption is given by the heating value of the fuel, Hi , mbed; ox
and the fuel power, Pfuel : hbed; ox = ; (9a)
(1 − )Aox
Pfuel mbed; red
ṁfuel = : (3) hbed; red = : (9b)
Hi (1 − )Ared
The amount of oxygen needed for oxidation of the fuel The oxygen ratio, Ro , is a property of the carrier and
is then signi@es the mass fraction of oxygen in the carrier (in its
ṁfuel fully oxidized state)
ṁo = MO2 Sr ; (4)
Mfuel Mf:ox − Mf:red
Ro = ; (10a)
where Sr is the stoichiometric ratio for the reaction be- Mf:ox
tween fuel and oxygen, and Mi denotes the molar mass of where Mf:red is the molar mass of the fully reduced carrier,
species i. The air ratio, , is given by the volume fraction and Mf:ox is the molar mass of the fully oxidized carrier.
of oxygen in the air from the oxidizer, xO2 ;ex The degree of oxidation, X , henceforth called conversion,
0:21(1 − xO2 ;ex ) is then the actual mass of oxygen divided by the mass of
= ; (5) oxygen when fully oxidized
0:21 − xO2 ;ex
Mactual − Mf:red
where 0.21 is the volume fraction of oxygen in air. (In X= ; (11a)
Mf:ox − Mf:red
contrast to normal combustion, the gas ow leaving the
oxidizer does not contain CO2 and water, which explains where Mactual is the actual molar mass of the carrier in
the presence of the xO2 ;ex in the numerator.) The air mass its partially oxidized state. The solids in the two reactors
ow is then are assumed to be well stirred, and the conversion of the
ṁo solids is therefore equal to the conversion of the solid
ṁair = ; (6) ows leaving these reactors. Since oxygen is transferred
0:233
from oxidizer to the reducer, the average conversion is
where 0.233 is the mass fraction of oxygen in air. The higher in the oxidizer, Xox , than in the reducer, Xred , and
uidizing velocity in the oxidizer is the di8erence in conversion, VX , is
ṁair va; air Tox VX = Xox − Xred : (11b)
uox = ; (7a)
Aox Ta
The capacity of the carrier, C, de@ned as the ratio of a
where va; i is the speci@c volume of gas i at ambient tem- fractional mass increase to an increase in conversion, is
perature, Aox is the cross-section area of the oxidizer, Tox a measure of how much oxygen the carrier is able to
is the temperature in the oxidizer and Ta is the ambient transfer for a given change in conversion
temperature. Similarly the uidizing velocity in the re-
ducer is 1 dm
C= ; (12)
m dX
ṁfuel va; fuel Rr Tred
ured = ; (7b) where m is the mass. For a conversion, X , equal to unity
Ared Ta
the capacity, C, is equal to Ro , but the capacity increases at
where the subscript red denotes the conditions in the lower conversions. After some manipulation the capacity
reducer and Rr is the recirculation ratio, i.e. ratio of the can be expressed as a function of the conversion for the
total ow of gas supplied to reducer and the fuel ow. oxidizer and the reducer
The uidizing velocity is based on the unconverted fuel. Ro
If the fuel is methane each fuel molecule is converted Cox = ; (13a)
1 − (1 − Xox )Ro
to three molecules, and accordingly the actual uidizing
velocity increases with a factor of three as the reaction Ro
Cred = : (13b)
proceeds. The amounts of bed material in the oxidizer 1 − (1 − Xred )Ro
3106 A. Lyngfelt et al. / Chemical Engineering Science 56 (2001) 3101–3113
The required conversion rates in the oxidizer and the (ii) the amounts of H2 , hydrocarbons and non-com-
reducer then become bustible gases in the recirculated gas can be neglected
compared to CH4 and CO.
d X ṁo
rox = = ; (14a)
d t ox Cox mbed; ox The conversion of the gas in the oxidizer is simply
d X ṁo 1
rred = = : (14b) $ox = : (20b)
d t red Cred mbed; red
The fan power, Pbed , necessary to overcome the pressure The present work discusses two oxygen carriers, Fe2 O3
drop, Vp, of the uidized beds is given by and NiO. In the case of Fe2 O3 , possible reduced forms
(n−1)=n include Fe3 O4 , FeO and Fe. The de@nition used for the
n p1 va ṁi p1 + Vp conversion of Fe2 O3 here is based on a full conversion to
Pbed = − 1 ; (15)
(n − 1) "fan p1 Fe, and the oxygen ratio is
MFe2 O3 − 2MFe
where n is the isentropic coeLcient, p1 is the inlet pres- Ro; Fe = ; (10b)
sure, here assumed to be ambient, ṁi is the mass ow of MFe2 O3
either air or fuel, and "fan is the eLciency of the fan. ’p , where Mi is the molar mass of species i. Similarly the
the fractional fan power to overcome the pressure drop carrier ratio of NiO is
is obtained as
MNiO − MNi
Pbed Ro; Ni = : (10c)
’p = : (16) MNiO
Pfuel
The power consumed by the fans is only partially lost,
since it is converted to thermal energy in the gas ows 6. Design criteria
entering the process. For this reason, the loss in eLciency
due to pressure drop, ’loss , depends on the overall eL- Design criteria were chosen for the layout of an at-
ciency of the power process, "tot : mospheric boiler with a power of 10 MW. Such a boiler
is suitable for the demonstration of the technology and
’loss = ’p (1 − "tot ): (17) for obtaining data and experience of the process in a
semi-commercial scale. The boiler could be used for heat
The mass ow of entrained solids from the oxidizer to production, for district heating or for industrial process
the reducer, ṁsol , is related to the di8erence in conversion steam. As seen in Fig. 2, the design resembles that of a
ṁo circulating uidized bed (CFB) boiler for combustion of
ṁsol = : (18a) solid fuels, thus using elements of proven technology. The
VXCox
boiler, when a demonstration=research programme is @n-
The mass ow of solids returned from the reducer to the ished, can therefore be converted to a conventional CFB
oxidizer is somewhat smaller, because of the oxygen used boiler, or conversely, an existing CFB boiler could be
in the reducer, converted to a CLC boiler. A full optimization is prema-
ture at this state of knowledge. Instead design data were
ṁsol; ret = ṁsol − ṁo : (18b)
chosen according to the following considerations, which
The conversion of the fuel, here CH4 , to CO2 , henceforth to some extent rely on experience from CFB boilers:
termed the gas yield, is
• The fuel is methane, since natural gas is a suitable fuel
xCO2
$red = ; (19) for a @rst application of the process.
xCO2 + xCH4 + xCO • The air ratio aims at keeping down the gas ow in the
where xi is the measured volume fraction of species i reactor, and yet have a suLcient oxygen concentration
in the product gas. The gas recirculation ratio, Rr , is the for high conversion rate of the solids.
actual amount of gas that has to be compressed after • The oxidizer cross-sectional area gives a similar
removal of the water divided by the CO2 produced. The gas velocity as in a CFB combustor, and the same
relation between gas yield and recirculation ratio is then cross-sectional area was chosen for the reducer, al-
though the gas velocity becomes smaller.
1
$red = (20a) • The temperatures reect the level where most of the
Rr experimental data are available at present. In the cal-
if the following assumptions are made: culations the choice of temperature only a8ects the gas
velocities. The temperature in the air reactor is some-
(i) gas losses from the recirculation loop can be ne- what higher because of the temperature fall in the fuel
glected except for the CO2 separated, reactor.
A. Lyngfelt et al. / Chemical Engineering Science 56 (2001) 3101–3113 3107
Table 3 version rates of the oxygen carriers were 9%= min in the
Chosen or assumed parameter values oxidizer and 3%= min in the reducer and the required mass
Item Symbol Value Unit ow of solids between the two uidized beds is approx-
imately 50 kg= m2 s. These results are realistic judging
Power (fuel) Pfuel 10 MW
Flue gas oxygen xO2 ;ex 0.04 dimensionless
from the experimental data, as will be discussed below.
Oxidizer cross section Aox 2.5 m2 The reactivity of the carrier can be expressed in terms
Reducer cross section Ared 2.5 m2 of conversion rate. The conversion rates, rox and rred ,
Temperature oxidizer Tox 1243 K are in turn linked to the gas yields, $ox and $red , and to
Temperature reducer Tred 1223 K the conversion di8erence of the solids, VX . A lowered
Recirculation ratio Rr 1.05 dimensionless
Pressure drop, oxidizer Vpox 7 kPa
gas yield or conversion di8erence raises the conversion
Pressure drop, reducer Vpred 20 kPa rates. All three parameters are important input for design,
Voidage in bed 0.6 dimensionless as shown in Fig. 3, and the interdependence of these
Fan eLciency "fan 0.9 dimensionless parameters is important in the optimization of the process.
Ambient pressure p1 101 kPa Therefore, laboratory studies of oxygen carrier reactivity
Carrier properties
Density of oxygen carrier 5000 kg= m3
should not only show the conversion rate, but also how
Oxygen ratio Ro 0.3 dimensionless the rate depends on gas yield and conversion di8erence
Conversion in oxidizer Xox 0.98 dimensionless of solids.
Conversion di8erence VX 0.02 dimensionless
7.1. Comparison to experimental data on reactivity
Table 4
Constants Conversion rates of 200 m natural hematite particles
Item Symbol Value Unit were studied by Mattisson et al. (2000b). The particles
were cyclically exposed to air and methane in a @xed-bed
Lower heating value Hi 50 MJ=kg
Stoichiometric ratio Sr 2 dimensionless
reactor, thus simulating the environment of particles mov-
Ambient temperature Ta 293 K ing between the air and fuel reactor. The oxidation rates
Molar mass air Mair 29 kg=kmol were high, signi@cantly above the desired 9%= min. The
Molar mass fuel Mfuel 16 kg=kmol reduction rates were obtained as a function of gas con-
Molar mass oxygen M O2 32 kg=kmol version and conversion range, see Fig. 4. A 95% yield of
Speci@c volumea of air va; air 0.841 m3 = kg
Speci@c volumea of fuel va; fuel 1.50 m3 = kg
gas was obtained at a conversion rate of 3–4%= min and
Gravity acceleration g 9.81 m2 = s with a conversion di8erence of 2%. Consequently, the
Isentropic coeLcient n 1.4 dimensionless gas yield and conversion rates of Table 5 are ful@lled. In
a At
a later study much higher reduction rates were obtained
ambient temperature.
with synthetic particles of Fe2 O3 and Al2 O3 (Mattisson et
al., 2000a).
• The recirculation ratio is low to minimize the power The rate data in Table 5 were derived for Fe2 O3 , but
needed for compression of CO2 . most literature data are given on NiO, normally in com-
• The pressure drop in the oxidizer is similar to that bination with an inert binder material. For comparison,
of CFB combustors. The pressure drop in the reducer desired conversion rates were also determined for a car-
is higher than in the oxidizer, in view of the lower rier composed of 60% NiO, and 40% inert material. The
conversion rates of the reduction. assumptions and the results are given in Table 6. A higher
• The conversion di8erence is small, which will give a conversion rate is needed for the NiO=inert material than
large recirculation rate of solids. With a large recircu- for Fe2 O3 since the oxygen ratio is lower for NiO=Ni, and
lation rate the temperature di8erence between the re- this oxygen ratio is further reduced by the presence of
actors becomes small. inert material, cf. Eqs. (10); (13) and (14).
• The oxygen carrier is hematite, but a comparison is The reactivity data in Table 2 show approximate con-
made with data on NiO=YSZ. The particle size is version rates for di8erent VX , probably obtained at low
200 m, although this is not used in the calculations. gas conversions. Assuming a @rst-order reaction, the con-
version rate is proportional to the concentration of the
The design values are listed in Table 3. The constants reactant. The e8ective concentration of the reactant gas
used are given in Table 4. can be approximated by the logarithmic mean value
xin − xout
xave = ; (21)
ln(xin =xout )
7. Results and discussion
where xin and xout are the volume fractions of the reactant
Eqs. (3)–(20) together with the values of Tables 3 and gas entering and leaving the reactor. In the case of air,
4 give the results presented in Table 5. The required con- the assumed inlet and outlet volume fractions are 0.21
3108 A. Lyngfelt et al. / Chemical Engineering Science 56 (2001) 3101–3113
Table 5
Parameter values derived from the basic assumptions in Table 3
Fig. 3. The basic relations between carrier reactivity and design input data.
and 0.04, yielding a logarithmic mean of 0.1. This means fraction of CH4 of 0.33 (Jin et al., 1999). Here the rate was
that the experimental conversion rates probably should 15%= min, Table 2, which, if corrected for partial pressure,
be reduced by a factor of two to allow a comparison with corresponds to about 10%= min, to be compared to the
the desired rates in Table 5. The oxidation rates of Jin et desired 7%= min, Table 5. Again the laboratory data were
al. (1998, 1999), about 20%= min, Table 2, therefore cor- obtained for large particles and much higher conversion
respond to approximately 10%= min, which is about half rates were found for small particles, e.g. 210%= min, see
of the desired 21%= min, see Table 6. These experimen- Table 2.
tal values were, however, obtained for large particles and Thus, the literature data support the conclusion that
smaller particles show much higher conversion rates, e.g. oxygen carriers can be found that ful@l the conversion
90%= min, as seen in Table 2. rates needed for the proposed design.
The concentration of the reactant gas leaving the fuel
reactor can be derived from the assumed recirculation 7.1.1. Bed mass and bed dimensions
ratio, 1.05, yielding a logarithmic mean volume fraction The needed bed mass, i.e. the bed volume (cross-section
of methane of 0.24. The best data for comparison are area × bed height), is determined by the conversion
those with a 1:2 mixture of CH4 and H2 O, i.e. a volume rate, Eq. (14a) and (14b). A higher bed raises the power
A. Lyngfelt et al. / Chemical Engineering Science 56 (2001) 3101–3113 3109
and storage (pp. 25 –35). Chalmers University of Technology and Nakano, Y., Iwamoto, S., Maeda, T., Ishida, M., & Akehata, T.
University of Gothenburg, GPoteborg, October 22. (available on (1986). Characteristics of reduction and oxidization cyclic process
http:==www.entek.chalmers.se= ∼anly=symp=sympco2.html) by use of Fe2 O3 medium. Iron & Steel Journal of Japan, 72,
Mattisson, T., Lyngfelt, A., & Cho, P. (2000b). The use of iron oxide 1521–1527.
as an oxygen carrier in chemical-looping combustion of methane Richter, H., & Knoche, K. (1983). Reversibility of combustion
with inherent separation of CO2 . Submitted for publication. processes. ACS Symposium Series, 235, 71–86.
Mattisson, T., Lyngfelt, A., & Cho, P. (2000a). Possibility of Riemer, P. (1998). Greenhouse gas mitigation technologies. An
using iron oxide as an oxygen carrier for combustion of overview of the CO2 capture, storage and future activities of the
methane with removal of CO2 — application of chemical-looping IEA greenhouse gas R& D programme. The IEA Greenhouse Gas
combustion. Fifth international conference on greenhouse gas R&D Programme: http:==www.ieagreen.org.uk=paper2.htm
control technologies, Cairns, Australia, 13–16 August (available
on http:==www.entek.chalmers.se= ∼anly=co2=ghgt5.pdf).