Beruflich Dokumente
Kultur Dokumente
www.elsevier.com/locate/ijhydene
Abstract
A high-temperature solar reactor has been developed for co-producing hydrogen-rich gas and high-grade carbon black (CB)
from concentrated solar energy and methane. The approach is based on a single-step thermal decomposition (pyrolysis) of
methane without catalysts and without emitting carbon dioxide since solid carbon is sequestered.
In the tested reactor, a graphite nozzle absorbs concentrated solar radiation provided by a solar furnace. The heat is then
transferred to the reactive flow. The experimental setup, first test results, and effect of operating conditions are described in
this paper.
The conversion of methane was strongly dependant on the solar furnace power input, on the geometry of the graphite
nozzle, on gas flow rates, and on the ratio of inert gas-to-reactive gas. CB was recovered in the carbon trap, and maximum
chemical conversion of methane-to-hydrogen and CB was 95%, but typical conversion was in the range 30–90%.
䉷 2004 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved.
Keywords: Hydrogen; Production; Solar energy; Thermal cracking; Methane; Carbon black
0360-3199/$30.00 䉷 2004 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijhydene.2004.09.006
844 S. Abanades, G. Flamant / International Journal of Hydrogen Energy 30 (2005) 843 – 853
Nomenclature
C mean flux concentration ratio QAr flow rate of argon in the inlet gas, Ln / min, Ln :
0
FCH molar flow rate of methane in the inlet gas, liter at normal conditions, 273 K, 1 atm
4
mol/s QCH4 flow rate of methane in the inlet gas, mLn / min
FCH4 molar flow rate of methane in the off-gas, mol/s Treactor temperature of the gas in the reaction zone, K
FH2 molar flow rate of hydrogen in the off-gas, TW wall temperature of the reactor, K
mol/s tr residence time, s
FAr molar flow rate of argon, mol/s X chemical conversion
H enthalpy change of the reaction, J mol−1 yCH4 mole fraction of CH4 in the off-gas
I direct solar flux density, W m−2 yH2 mole fraction of H2 in the off-gas
Qsolar solar power input, W th thermochemical efficiency
has zero CO2 emission: carbon as opposed to CO2 is se- Thermodynamics of methane decomposition
2
questered and marketable CB is synthesized; (d) moreover,
fossil fuels are saved with respect to the production of H2 Solid carbon, C
and CB by conventional methods. 1.5
This solar route saves the process energy and the CO2 Methane, CH4
1 Molecular
by classical methods (methane steam reforming and partial hydrogen, H2
up to 9 kW and an indirect heating aerosol flow reactor are • geometry of the graphite nozzle (opaque heat transfer
reported in [5,18]. Methane conversion approaching 70% cavity that absorbs the solar radiation) designed to per-
was obtained at 2000 K for a residence time in the order of form the solar thermal decomposition of CH4 .
0.1 s [5]. The effect of the reactor wall temperature (TW ) and
the residence time (tr ) on the conversion was studied in [18]: The test results include:
approximately 90% conversion of methane was obtained at
TW = 2133 K (power level: 9 kW) and tr = 0.01 s. At 1923 K • mole fraction of methane and hydrogen in the off-gas,
(power level: 6 kW), the conversion reduced to 35%. • decomposition yield of the reacting gas measured with
The ETH’s solar chemical reactor [14] features a vortex respect to CH4 , and flow rate of produced H2 .
flow of CH4 confined to a cavity-receiver and laden with
carbon particles. The highest conversion occurs when carbon
particles are fed with the gas. With a 5 kW reactor prototype, 2. Experimental setup and method
maximum chemical conversion of CH4 -to-H2 and C(gr) was
67% at 1600 K and 1 bar. The schematic diagram of the experimental system is
The solar reactor prototype must be designed accounting shown in Fig. 2. The process that performs methane crack-
for several considerations: choose the technical options for ing is composed of the following subsystems: concentrating
(1) injecting the reactant; (2) heating the reacting gas and system (flat heliostat + parabolic mirror), solar reactor (ab-
carbon nano-particles by solar radiation; (3) protecting the sorbing radiation), filtering system for gas and solid sepa-
optical window against carbon particles deposition; (4) pre- ration, and on-line gas diagnostic system for measuring ex-
venting clogging by carbon particles inside the reactor. haust gas composition. The solar reactor design integrates
In this study, a lab-scale solar reactor has been designed the following three main functions: reactant heating, prod-
and constructed in order to study the production of hydro- ucts cooling (quenching), and solid–gas separation.
gen and CB from thermal cracking of methane. First, we The solar reactor body was made of AISI 316L stainless
chose a heat transfer wall that consists in a cylindrical cav- steel (88 mm i.d., 110 mm o.d., and 150 mm long), and was
ity directly exposed to the high-flux solar radiation. The gas water-cooled. Concentrated solar radiation entered inside
flows directly in the irradiated zone. Once the methane de- the reactor through a PyrexTM window. The inside part of
composition has begun, the carbon particles that are formed the reactor was made of a graphite inner tube 17 mm o.d.
also serve simultaneously as radiant absorbers and as nu- and 61 mm long. This inner tube, so-called “nozzle”, was
cleation sites for the heterogeneous decomposition reaction surrounded by an other graphite tube 22 mm i.d. and 61 mm
[14]. Consequently, the mechanisms of heat transfer to the long. The annular region between the two tubes, not yet used,
reactive flow injected in the cavity are mainly forced con- will permit to test another reactor configuration (the indirect
vection between the gas stream and the cavity walls, and heating mode will be studied by injecting methane between
direct particle heating by radiative heat transfer (solar spec- the two graphite tubes). The inner part of the reactor was
trum). The IR radiation emitted from the hot cavity walls insulated by a surrounding layer of carbon felt. The reaction
may also contribute to heat the gas (CH4 absorbs in the IR occurred in the graphite nozzle heated by concentrated solar
spectrum) and the particles. energy. Various shapes of this nozzle were tested.
Experiments were performed at different scales. Two The carrier gas, argon, was injected either at the entrance
parabolic concentrators were tested: 1.5 m-diameter and of the graphite nozzle, or at the top of the PyrexTM window
2 m-diameter, with vertical axis solar facilities existing at mainly to prevent the deposition of carbon particles on the
the Odeillo solar furnace laboratory. The solar furnace pro- windows, which could cause its breakage. Argon was also
vided a vertical axis concentrated solar beam focused on fed into the reactor in order to provide an inert atmosphere
the target. The target was made of graphite and was nozzle- to prevent oxidation of the graphite tube. The flow rates of
shaped. The reactant was heated at the nozzle wall where argon and methane were measured and controlled by mass
a part decomposed and then flew down to a cooling zone. flow controllers (MFC, Brooks Instruments model 5850 S).
After this zone, a filtering system separated the solid–gas An adapted filtering media was installed at the exit of the
mixture. reactor after the mixing/cooling zone to separate the carbon
This paper focuses on the development of the solar pro- nanoparticles from the gas–solid flow. The main part of the
cess for thermal splitting of methane. The design and con- solid product, CB, was recovered in this carbon trap, the
struction of the reactor prototype are described, and experi- other part deposited on the reactor walls. The pressure was
mental results are presented. The main parameters of these maintained about 40 kPa below the atmospheric pressure in
first tests include: order to achieve a continuous gas flow through the filtering
device. This system was implemented to maintain a contin-
• solar furnace power input/Size of the solar concentrator uous removal of carbon particles, and to prevent clogging
(parabola), by carbon particles inside the reactor.
• flow rates of Ar and CH4 , and ratio between buffer gas The temperature was measured at various locations (lat-
(Ar) and reacting gas (CH4 ) mass flow rates, eral zone T1 , mixing/cooling zone T2 , filter T3 , filter exit T4 )
846 S. Abanades, G. Flamant / International Journal of Hydrogen Energy 30 (2005) 843 – 853
MFC
Fluorine Data
window Ar +CH4 acquisition
workstation
PyrexTM glass
T wall (optical pyrometer)
Gas sample
T2 mixing/cooling Water
zone
Pressurized To vent
air Primary
vacuum
pump
Fig. 2. Schematic configuration of the solar experimental setup for the thermal cracking of CH4 .
by K-thermocouples. The mean temperature of the graphite absorb the beam-down radiation source. It was preheated to
tube TW was measured by an optical pyrometer (wavelength: the maximum temperature that can be reached by the sys-
5.12 m). The measurement was made through a vertically tem which depends on the concentrating system and on the
centered port made of fluorine (CaF2 ) located at the top of direct solar irradiation. The duration to reach steady state
the window (Fig. 2), the diameter of the measured area was was about 10–15 min. The solar furnace consisted of a sun-
about 6 mm. tracking heliostat in-axis with a parabolic solar concentra-
In addition, the off-gas composition and species mole tor. Preliminary experiments were performed on a 1.5 m-
fraction in exhaust gases were measured on-line by gas chro- diameter concentrator, then a 2 m-diameter concentrator was
matography (GC) throughout an experimental run. The gas used to reach higher temperature and higher conversion.
outlet was connected to a nearby micro GC (Varian CP 4900) At steady state, CH4 was injected into the reactor via
equipped with thermal conductivity detector (TCD) and two either the inlet port located at the top of the reactor also
columns. The gas to be analyzed was aspirated through a set used to inject Ar (top inlet port), or through a radial port
including a primary sampler (peristaltic pump) and a sec- located at the entrance of the graphite nozzle. In the latter
ondary sampler (internal pump of the gas chromatograph). case, auxiliary Ar (0.1 Ln / min) was also fed with CH4 via
The micro GC was equipped with 2 channels: Channel 1 the radial inlet port.
(MolSieve 5A PLOT 10M Backflush) for H2 , N2 , O2 , CO,
CH4 ; Channel 2 (PoraPLOT U 10M Backflush) for hydro-
carbons such as CH4 , C2 H6 , C2 H4 , C2 H2 , CO2 , and H2 . 3. Experimental conditions and results
Each channel was a separate GC with pneumatics, injec-
tor, column, and micro-machined TCD. The carrier gas used Methane can be decomposed into hydrogen and CB ac-
for chromatography analysis was Argon. Since Ar was also cording to the simplified scheme shown in Eq. (1)
used during methane cracking experiments, the matrix ef- CH4 → C (solid) + 2H2 , H ◦ = 75.6 kJ/mol. (1)
fects were eliminated. A membrane separator (Genie Model
170), set at the entry of the GC, removed entrained liquid Methane decomposition reaction is a moderately endother-
and particulate in gas samples, thereby preventing contam- mic process. High reaction temperature and low reaction
ination or damage to the gas analyzer. pressure favor methane cracking reaction.
The acquisition of data was performed with an appropriate Different geometries of the nozzle target were tested, be-
Control/Data Handling Software (CP-Maître Elite) and all cause it governs gas-wall heat transfer, which is a key pa-
data were recorded and stored on a PC workstation. rameter for efficient methane decomposition. The different
During experiments, the reactor was first set at the focus cross sections of the tested graphite nozzles are represented
of the solar furnace under a flow of inert gas (Ar feeding). in Fig. 3. The influence of the aperture diameter and of the
Ar main flow rate was injected at the top inlet port of the cavity length was studied. The largest diameter of the noz-
reactor. The reactor was positioned vertically, facing up, to zle was 10 mm and the length of the cavity that absorbs so-
S. Abanades, G. Flamant / International Journal of Hydrogen Energy 30 (2005) 843 – 853 847
10 10 10 6
1
10 10 15
30
60
7 10
17 3
3 3 3
(a) (b) (c) (d) (e) (f) (g)
lar radiation varied from 10 to 30 mm (total length of the the measurement of the gas temperature in the hot reactive
nozzle: 60 mm). zone.
Experiments were carried out with various ratio of On-line monitoring consists in the measurement of the
methane to argon in the flow. In most experiments, Ar off-gas composition (H2 , CH4 , C2 H2 , C2 H4 , C2 H6 ). The
flow rate was 0.9–1 Ln / min (Ln means liter at normal composition is measured by gas chromatography (GC) as
conditions, 273 K, 1 atm) depending whether auxiliary Ar a function of time, analyzing a bypass stream of the pro-
was used or not. The flow rate of methane varied from duced gas at the reactor exit. As results, methane conver-
100 to 700 m Ln / min, thus CH4 mole fraction in the feed sion, off-gas composition, hydrogen productivity (flow rate
was in the range 9–44%. After preheating, the duration of of produced H2 ), and CB microstructure can be determined.
experiments with Ar–CH4 feeding was 15–30 min. The following presents the results related to the gas prod-
Under Ar feeding during preheating, the maximum tem- ucts analysis, CB will be characterized later to evaluate their
peratures reached were at steady state: T1 = 820 K, T2 = quality compared to the standard industrial production.
385 K, T3 = T4 = 293 K (ambient temperature). The injec- The chemical conversion, X, is defined as the molar ra-
tion of CH4 led to a decrease of temperatures and to a new tio of CH4 decomposed to CH4 in the feed (X = 1 −
equilibrium: T1 = 760 K, T2 = 360 K, T3 = T4 = 293 K. The FCH4 /FCH0 ). It is calculated assuming that secondary re-
4
temperature discrepancy between the upper part of the noz- actions (more particularly the reaction leading to acetylene)
zle and the mixing/cooling zone indicates that an important are negligible. In other terms, all the methane decomposed
temperature gradient exists along the nozzle, and that the is supposed to form hydrogen and carbon. Thus, X is cal-
cooling rate of the gas products is high (about 1500 K/s). culated from the mole fraction of methane in the off-gas
An efficient quenching avoids products recombination and which can be written as
thus leads to higher hydrogen yield. If the quenching is too 0 (1 − X)
rapid (thus the reaction time limited), intermediate products FCH
yCH4 = 4
0 (1 − X) + 2F 0 X
. (2)
(mainly acetylene) can be obtained from the pyrolysis of FAr + FCH CH4
4
methane [19].
The typical wall temperature of the reactor (upper zone) Actually, small quantities of pyrolytic hydrocarbons such
varied between 1300 and 1700 K, but measurements were as acetylene do form during high-temperature solar crack-
performed on the graphite cone surrounding the nozzle en- ing and are detected, but it does not affect significantly the
trance because of the too large measured area of the py- value of the chemical conversion measured with respect to
rometer (6 mm diameter). Thus, this temperature is not the methane since their amount in the off-gas is low in the range
real mean temperature of the wall in the reacting zone and of methane flow rates studied. Among the secondary species
the values are not reported in the results section. Hence, detected, acetylene is the most important and its mole frac-
a more accurate method for wall temperature measurement tion varies from 1% to 5% according to GC analysis. The
will be developed in the next step of the study. Moreover, mole fraction of ethylene is 0.1–0.4%, and the mole fraction
the temperature of the gas flow is probably lower than the of ethane is about 10 times less than that of ethylene.
wall temperature because of the short residence time of In order to determine closely the exiting gas flow rates of
the gas in the nozzle. Thus, a model is needed in order to each species and to obtain closure on a material balance, an
evaluate the gas temperature profile in the reaction zone. internal standard such as helium will be fed with the starting
It must account for transport mechanisms and chemical re- methane–argon mixture (at a mole fraction of about 1%) in
action. A type D thermocouple will be used to complete future experiments. Using this method, since helium will not
848 S. Abanades, G. Flamant / International Journal of Hydrogen Energy 30 (2005) 843 – 853
Table 1
Operating conditions and summary of results obtained during the solar experimental campaign with the 1.5 m-diameter concentrator
Run 1 2 3 4 5 6
Conditions
Direct solar flux density 950 950 1044 1044 1017 1017
(W m−2 )
Nozzle configuration b b e e d d
Ar main flow rate (Ln / min) 0.9 0.9 0.9 0.9 0.9 0.9
Auxiliary Ar radial flow rate 0.1 0.1 0.1 0.1 0.1 0.1
(Ln / min)
CH4 flow rate at top inlet port 0 0 0 0 0 0
(Ln / min)
CH4 radial flow rate (Ln / min) 0.1 0.3 0.1 0.3 0.1 0.2
Initial CH4 mole fraction 0.091 0.231 0.091 0.231 0.091 0.167
CH4 mole fraction in the off-gas, 0.0730–0.071 0.198 0.0780–0.075 0.189–0.175 0.076–0.071 0.128–0.123
yCH4
CH4 conversion (%) 18.1–20.8 11.6 13–15.9 15.2–20.5 15–20.5 20.6–23.4
H2 mole fraction in the off-gas, 0.026–0.029 0.091 0.016–0.018 0.040–0.048 0.016–0.029 0.050–0.054
y H2
Solar power input Qsolar (W) 400 400 440 440 428 428
Thermochemical efficiency th (%) 0.7 1.1 0.5 1.8 0.6 1.4
15%
During the experimental study performed with a 1.5 m- Tables 2 and 3 list the results obtained with a 2 m-diameter
diameter solar concentrator, the maximum value of methane concentrator and the corresponding operating conditions.
conversion is 23% whatever the design of the nozzle. In- In Run 1, the nozzle target position is moved from 10 mm
deed, it is limited by the temperature in the reacting zone. up to 3 mm beneath the focus of the solar furnace. When
S. Abanades, G. Flamant / International Journal of Hydrogen Energy 30 (2005) 843 – 853 849
Table 2
Operating conditions and summary of results obtained during the solar experimental campaign with the 2 m-diameter concentrator
Run 1 2 3 4 5 6 7
Conditions
Direct solar flux density (W m−2 ) 960 1000 1000 980 980 980 980
Nozzle configuration d c c a a a a
Ar main flow rate (Ln / min) 0.9 0.9 0.9 0.9 0.9 0.9 0.9
Auxiliary Ar radial flow rate (Ln / min) 0.1 0.1 0.1 0.1 0.1 0.1 0.1
CH4 flow rate at top inlet port (Ln / min) 0.1 0.1 0.2 0.1 0.2 0.4 0.6
CH4 radial flow rate (Ln / min) 0 0 0 0 0 0 0
Initial CH4 mole fraction 0.091 0.091 0.167 0.091 0.167 0.286 0.375
CH4 mole fraction in the off-gas, yCH4 0.068–0.051 0.035–0.030 0.059 0.037–0.032 0.061–0.057 0.108–0.106 0.151-0.117
CH4 conversion (%) 23.5–41.1 59.8–65.2 61.0 57.6–62.7 59.7–62.4 56.2–56.7 51.8–61.5
H2 mole fraction in the off-gas, yH2 0.037–0.064 0.091–0.099 0.163 0.083–0.096 0.163–0.187 0.280–0.288 0.308–0.340
Solar power input Qsolar (W) 772 805 805 788 788 788 788
Thermochemical efficiency th (%) 0.4–0.7 1 1.9 1 2 3.6 5.9
the nozzle target is gradually moved up towards the focus decrease in the reacting zone, which explains the conversion
of the solar furnace, the peak concentration increases and decrease.
the conversion of methane increases too from 23% to 41% Fig. 7 shows the variation of CH4 conversion as a func-
(Run 1) because the temperature in the nozzle obviously tion of the geometry of the nozzle and the operating condi-
increases. tions (CH4 flow rate). The most efficient target for methane
For nozzle types (c) and (a), the conversion of methane is cracking is nozzle type (g) because it exhibits the largest
about 60%, and it is neither significantly influenced by the available specific reaction surfaces. The higher the flow rate
flow rate of methane, nor by the duration of the run (Runs of methane, the lower the conversion for each target tested.
2 to 7).
The following results presented in Figs. 5–7 are listed in 3.3. Thermal efficiencies
Table 3 (Ar and CH4 are injected via the top inlet port in
runs 8–16). The thermochemical efficiency th of the solar reactor
Fig. 5 shows the results obtained for a type (g) graphite has been evaluated for the experiment with the 2 m-diameter
target: a graphite wall (length 15 mm, thickness: 1 mm) was concentrator. It represents the amount of solar energy which
inserted in the center of the cylindrical nozzle in order to is stored chemically into hydrogen. This efficiency is defined
increase the heat exchange area between the reacting gas and as [14]
the target. In this case, the conversion of methane reaches
0 XH
th = FCH
95%. The conversion of methane decreases with the Ar flow
4 Reactants(298 K)→Products(Treactor ) /Qsolar ,
rate for a constant flow rate of methane (100 mLn / min), as (3)
illustrated in Fig. 5. Besides, for a given flow rate of Ar,
the conversion of methane decreases with time. The off- where FCH0 is the inlet molar flow rate of methane (mol/s),
4
gas composition is analyzed about every 2 or 3 min (time X is the chemical conversion, H is the enthalpy change
duration between two consecutive vertical plots) since the of the reaction when the reactants are fed at 298 K and the
sampling time lasts 60 s, and the run time acquisition for products are obtained at Treactor (170 kJ/mole at Treactor =
one chromatogram (both channels) lasts also 60 s. 1500 K), and Qsolar is the solar power input (W m−2 ).
The conversion of methane decreases when the CH4 flow Qsolar is estimated by integrating the profile of solar flux
rate increases for a constant Ar flow rate, i.e. when the density obtained at the focal plane (given in appendix).
mole fraction and the flow rate of hydrogen in the product The thermochemical efficiency is in the range 1.3–4.5%
stream increases (Fig. 6). The same effect of the CH4 flow for a nozzle type (g) (i.e. an aperture of 10 mm with a
rate was observed by Dahl et al. [18]: they found no appar- graphite slab inserted in the center of the nozzle) that ex-
ent conversion at high initial methane flow rates. Similarly, hibits the highest chemical conversion, and assuming that
Hirsch and Steinfeld [14] obtained the maximum CH4 con- the mean gas temperature in the nozzle is 1500 K. If the sen-
version (67%) at 1600 K with the lowest CH4 mole frac- sible heat absorbed by the inert gas is taken into account,
tion in the feed (11%) and with carbon particle co-feeding. the thermal efficiency is higher (4–8%). Hirsch and Stein-
Thus, the lower the inlet flow rate of methane and argon, feld [14] report efficiencies in the range 3–15% and 7–16%
the higher the chemical conversion of methane to hydrogen. without and with accounting for the sensible heat of Ar, re-
Actually, a global flow rate increase leads to a temperature spectively. These values indicate that the main part of solar
850 S. Abanades, G. Flamant / International Journal of Hydrogen Energy 30 (2005) 843 – 853
0.149
0.177
100%
1010
0.25
35.2
813
1.2
0.4
0.5
Flow rate of methane
16
f
100 mLn/min
Nozzle type (g)
Conversion of methane, X
0.049– 0.062
90%
0.052–0.044
31.3–41.1
Time
increase
0.077
1010
813
1.2
0.1
0.6
15
80%
f
0.119– 0.121
0.218–0.219
45.7–47.0
70%
1010
0.25
813
1.2
0.4
2.9
14
60%
0.8 0.9 1 1.1 1.2 1.3
0.064–0.063
0.141–0.143
Flow rate of Ar, QAr (Ln/min)
51.7–52.2
813
1.2
0.2
1.6
13
0.031–0.027
0.088–0.092
80% 400
Operating conditions and results about the influence of gas flow rates and nozzle configuration (2 m-diameter concentrator)
57.6–62.9
0.077
1010
813
Conversion of methane, X
1.2
0.1
60% 300
a
(mLn/min)
0.103–0.060
0.234–0.283
53.2–71.5
40% 200
1010
0.25
813
1.2
0.4
4.5
11
20% 100
Flow rate of Ar 1.2 Ln/min
0.019–0.010
0.094–0.117
0% 0
0.077
813
1.2
0.1
1.3
10
0.020–0.010
0.069–0.094
805
0.1
1.4
9
g
1
diation instead of being used to heat the gas and carry out
0.155–0.161
797
0.9
0.1
1.5
g
8
80% of CH4 conversion with time (Fig. 5). The challenge con-
sists in designing a reactor suitable for decomposition of
60% methane with production of hydrogen-rich gas, continuous
withdrawal of elemental CBs, and negligible pyrocarbon de-
40% position.
In the next step of the study, the wall temperature in the
reacting zone will be measured more precisely in order to
20%
understand the mechanisms leading to the carbon nanopar-
ticle formation in the gas phase and on the graphite walls.
0%
The method to be implemented will involve a solar blind
CH4 100 mLn/min
Type (g)
Type (a)
Type (a)
Type (f)
Type (f)
points out a small cavity (diameter: 3 mm, length: 15 mm)
pierced along the nozzle close to the heat transfer wall, and
that exhibits a blackbody behavior. The chemical yield (Y )
will be evaluated with respect to H2 , and calculated as the
Fig. 7. Effect of the geometry of the nozzle and CH4 flow rate on
ratio of H2 molar flow rate in the off-gas (determined from
CH4 conversion.
GC measurement with an internal standard such as He) to
the theoretically maximum H2 molar flow rate when the re-
action reaches completion (Y = FH2 /2FCH 0 ). In addition to
4000 4
the mole fractions, the quantity (flow rates) of secondary hy-
Pyrocarbon
deposit drocarbons species from methane (especially C2 H2 , C2 H4 ,
3000 and C2 H6 ) will be determined with accuracy, and experi-
mental conditions necessary to lower their formation will be
Counts/s
defined.
2000
Carbon Black
powder
1000 4. Conclusions
conditions defined, the experimental results showed that the is determined by integration of the flux density
nearly complete dissociation of methane is possible with R
this reactor configuration. In the next step of the study, the Qsolar = 2I rC(r) dr. (6)
performance of the solar reactor will be optimized with re- 0
spect to reactor thermal efficiency and maximal conversion
The energy Qsolar intercepted by a 10 mm-diameter disk is
of methane to hydrogen.
about 805 W (I = 1000 W m−2 ; Cmax = 16 000).
Acknowledgements References
This study was financially supported by the French [1] Steinberg M. Fossil fuel decarbonization technology for
CNRS, and Minister of Research and New Technologies in mitigating global warming. Int J Hydrogen Energy
1999;24:771–7.
the frame of the Energy program. Authors thank N. Boul-
[2] Muradov NZ. How to produce hydrogen from fossil fuels
let (PROMES-CNRS) for the technical support during the
without carbon dioxide emission. Int J Hydrogen Energy
construction of the solar reactor. 1993;18:211–5.
[3] Fulcheri L, Schwob Y. From methane to hydrogen, carbon
and water. Int J Hydrogen Energy 1995;20(3):197–202.
[4] Barreto L, Makihira A, Riahi K. The hydrogen economy in
Appendix
the 21th century: a sustainable development scenario. Int J
Hydrogen Energy 2003;28:267–84.
The concentrator is a parabolic mirror of 2 m-diameter, [5] Dahl JK, Buechler KJ, Finley R, Stanislaus T, Weimer AW,
85 cm focal distance, and 120◦ aperture. The solar power Lewandowski A, Bingham C, Smeets A, Schneider A. Rapid
input Qsolar intercepted by the target is given by solar-thermal dissociation of natural gas in an aerosol flow
reactor. Energy 2004;29:715–25.
Qsolar = I · A · C, (4) [6] Hirsch D, Epstein M, Steinfeld A. The solar thermal
decarbonization of natural gas. Int J Hydrogen Energy
2001;26:1023–33.
where C is the mean flux concentration ratio over a targeted [7] Cheynet B. Complex chemical equilibria calculation with the
area A at the focal plane, and I the incident normal beam THERMODATA system, Computer Software in Chemical and
insolation. The concentration profile can be fitted as Extractive Metallurgy. Proc Met Soc CIM 1988;11:31–44.
[8] Dahl JK, Barocas VH, Clough DE, Weimer AW. Intrinsic
C(r) = Cmax exp(−(r/g)2 ) (5) kinetics for rapid decomposition of methane in an aerosol
flow reactor. Int J Hydrogen Energy 2002;27:377–86.
[9] Trommer D, Hirsch D, Steinfeld A. Kinetic investigation of
with g = 5.1 mm (r in mm), and Cmax = maximum the thermal decomposition of CH4 by direct irradiation of
concentration. a vortex-flow laden with carbon particles. Int J Hydrogen
The shape of the flux density curve in the focal plane vs. Energy 2004;29:627–33.
the distance from the focal axis is represented in Fig. 9. The [10] Dahl JK, Tamburini J, Weimer AW, Lewandowski A, Pitts R,
energy received by a small disk (radius R) at the focus Bingham C. Solar-thermal processing of methane to produce
hydrogen and syngas. Energy Fuels 2001;15:1227–32.
[11] Olsvik O, Rokstad OA, Holmen A. Pyrolysis of methane in
the presence of hydrogen. Chem Eng Technol 1995;18:349–
1500
58.
[12] Billaud F, Guéret G, Weill J. Thermal decomposition of pure
methane at 1263 K. Experiments and mechanistic modelling.
Flux density (W/cm2)
1000
Thermochim Acta 1992;211:303–22.
[13] Back MH, Back RA. Thermal decomposition and reactions of
methane. In: Albright LF, Crynes BL, Corcoran WH., editors.
Pyrolysis: theory and industrial practice. New York: Academic
500 Press; 1983. p. 1–24.
[14] Hirsch D, Steinfeld A. Solar hydrogen production by thermal
decomposition of natural gas using a vortex-flow reactor. Int
J Hydrogen Energy 2004;29:47–55.
0 [15] Kogan A, Kogan M. The tornado flow configuration—an
-20 -15 -10 -5 0 5 10 15 20 effective method for screening of a solar reactor window. J
Distance from the focal axis (mm) Solar Energy Eng 2002;124:206–14.
[16] Kogan M, Kogan A. Production of hydrogen and carbon black
Fig. 9. Energy density repartition at the focus of the parabolic 2 m- by solar thermal methane splitting. I The unseeded reactor.
diameter solar concentrator vs. the distance from the focal axis. Int J Hydrogen Energy 2003;28:1187–98.
S. Abanades, G. Flamant / International Journal of Hydrogen Energy 30 (2005) 843 – 853 853
[17] Kogan A, Kogan M, Barak S. Production of hydrogen wall aerosol flow reactor. Int J Hydrogen Energy 2004;29:
and carbon by solar thermal methane splitting. II. Room 725–36.
temperature simulation tests of seeded solar reactor. Int J [19] Holmen A, Olsvik O, Rokstad OA. Pyrolysis of natural
Hydrogen Energy 2004;29:1227–36. gas: chemistry and process concepts. Fuel Process Technol
[18] Dahl JK, Buechler KJ, Weimer AW, Lewandowski A, 1995;42:249–67.
Bingham C. Solar-thermal dissociation of methane in a fluid-