Renewable Energy 222 (2024) 119952

Contents lists available at ScienceDirect

Renewable Energy
journal homepage: www.elsevier.com/locate/renene

Numerical study on a novel solar-thermal-reaction system for clean

hydrogen production of methanol-steam reforming
Ning Zhao a, Jiangjiang Wang a, *, Yuyang Tian a, Zibo Yao a, Suying Yan b
a
Hebei Key Laboratory of Low Carbon and High Efficiency Power Generation Technology, North China Electric Power University, Baoding 071003, Hebei, China
b
School of Energy and Power Engineering, Inner Mongolia University of Technology, Hohhot 010051, Inner Mongolia, China

A R T I C L E I N F O A B S T R A C T

Keywords: Renewable energy-driven hydrogen production methods offer a promising solution for mitigating energy crises.
Methanol steam reforming (MSR) The present study introduces a new solar-driven hydrogen production system that within methanol steam
Reaction kinetics reforming (MSR) process. The numerical analysis is carried out using a cylindrical Fresnel linear concentrator
Linear fresnel concentrator
coupled with a solar thermal reactor (STR). A comprehensive kinetic model of the MSR reaction was constructed
Solar thermal reactor (STR)
and combined with analytical methods based on Monte Carlo ray tracing methods and finite volume methods to
analyze the multi-physical properties involved in the reaction process. The results indicate that a novel solar-
driven MSR-STR system can generate a sufficient driving force after adjusting the concentrator structure and
relative position to match the energy demand of the MSR reaction. The accuracy of the model is compared with
the experimental values, and the average error is 7.8 %. Tracking error of the component should be maintained
below 1.5◦ to ensure the optical efficiency. The annual hydrogen production was analyzed and the levelized cost
of hydrogen is $1.77/kg. Finally, a comparison is presented between the proposed MSR-STR system and para­
bolic trough concentrator based MSR hydrogen production system to evaluate the feasibility of the former in
terms of the economics of implementation.

solar energy, sunlight is concentrated in a collector (also a reactor) to

1. Introduction raise the temperature to provide reaction power for thermochemical
heat absorption reactions. Methanol, known as “liquid sunlight,” has
The availability and sustainability of energy resources are crucial for many advantages over other liquid fuels, such as low boiling point,
the development of society [1]. In the face of the limited stock of fossil simple structure, easy storage and transportation, and high hydro­
energy and the accompanying environmental problems, the renewable gen/carbon ratio [15]. Reactors based on catalytic methanol-steam
energy is expected to be an effective means to escape the dilemma [2]. reforming (MSR) have been extensively studied [16–20]. Various ad­
Solar energy, an inexhaustible source of clean energy, needs to be vances have also studied the turbulent flow characteristics inside the
addressed for its volatility and intermittency. Hydrogen is the most thermochemical reactor [21,22] and the heat transfer performance [23].
promising clean energy carrier, with excellent energy release potential For solar-driven MSR reactions, key factors such as solar flux distribu­
and satisfactory energy density [3,4]. Hydrogen production from solar tion on the receiver surface [24,25], solar spectrum splitting, truncation,
energy for storage and utilization is being extensively studied, and the and distribution also have important implications for system design and
mainstream methods include fossil fuel reforming reactions [5,6], optimization [26,27].
electrolysis [7], and pyrolysis [8], and photocatalysis [9]. Hydrogen Matching different thermochemical hydrogen production conditions
production via solar thermochemical (STC) reactions has become a focus with corresponding solar concentrators is conducive to improving sys­
of scientific research and engineering applications as it shows renew­ tem stability and feasibility [28]. The solar dish reactor is compact and
able, sustainable, and low-cost potential in solar photothermal utiliza­ has a high concentration temperature, making it more suitable for
tion [10,11]. combining with methane steam reforming reactions (>1000 K). How­
Many solar-driven thermochemical technologies are available for H2 ever, methanol steam reforming process can be coupled with medium-
production, including steam reforming, partial oxidation, and self- and low-temperature concentrating solar collector modules, which can
heating reforming [12–14]. Based on the long-wave thermal effect of utilize solar energy to provide a heat source for the heat absorption

* Corresponding author.
E-mail address: wangjj@ncepu.edu.cn (J. Wang).

https://doi.org/10.1016/j.renene.2024.119952
Received 13 July 2023; Received in revised form 7 December 2023; Accepted 3 January 2024
Available online 4 January 2024
0960-1481/© 2024 Elsevier Ltd. All rights reserved.
N. Zhao et al. Renewable Energy 222 (2024) 119952

Nomenclature dgi inner glass tube diameter

dso outer steel tube diameter
Abbreviations dsi inner steel tube diameter
CPC compound parabolic concentrators ηo optical efficiency of the STR
CFD computational fluid dynamics ηt thermal efficiency of the STR
FVM finite volume methods ηc methanol conversion rate
LCFL linear cylindrical Fresnel lens n Materials’ refractive index
LFR linear Fresnel reflectors W width of the Fresnel lens with spike
LCOH levelized cost of hydrogen ρp density of the catalyst particles
MSR methanol steam reforming δg thickness of the glass cover
MCRT Monte Carlo ray tracing δl,max maximum thickness of the Fresnel lens
MSRR methanol steam reforming reaction δl,min minimum thickness of the Fresnel lens
MDR methanol decomposition reaction a aperture of the Fresnel lens
PDC parabolic dish concentrator L Active length of the solar thermal reactor
PTC parabolic trough concentrators SCO the CO selectivity
PTSR parabolic trough solar receiver
rWGSR reverse water-gas shift reaction Variables
STR solar thermal reactor c1F
S2S surface-to-surface c2F inertial loss term
SMR molar ratio between steam and methanol in inlet CTC total investment cost
WFR the ratio of catalyst weight to inlet methanol molar rate CO&M annual operation and maintenance cost
CMethanol
Parameters Dp average diameter of the catalyst particles
cp heat capacity εp porosity of the catalyst bed
λ thermal conductivity pi the partial pressure of species
ho enthalpy of species Nc grid number at the methanol reforming reactor surface
Ri the volumetric mass rate of creation of species nCH3OH, in inlet methanol flow rate
Re Reynolds number nCH3OH, out outlet methanol flow rate
Nu Nusselt number nCO2 molar flow rate of produced CO2
fm friction factor nCO molar flow rate of produced CO
fCR capital recovery factor β longitudinal incidence angle
qa,m average concentrating solar flux density γ tracking error angle, ◦
σq nonuniformity flux distribution f focus distance, m
dgo outer glass tube diameter

reaction while avoiding high-temperature catalyst deactivation. Cheng thermochemical catalysis with light-assisted thermal catalysis reactions
et al. designed an adequate parabolic trough solar receiver (PTSR) and increased the overall solar hydrogen production efficiency to 24.1 %
reactor and explored its potential for commercial scale-up through after Fresnel lens concentration [36]. Leutz et al. introduced an optimal
geometric parameter optimization [29]. Yu et al. filled the PTSR with convex non-imaging Fresnel lens was designed based on the edge ray
nanocatalysts to provide reaction heat through a significant photo­ principle [37]. Feng et al. reported the optical efficiency of the linear
thermal conversion effect [30]. Zhu et al. experimentally evaluated the cylindrical Fresnel lens (LCFL)at different incidence angles based on
comprehensive performance of the PV-PTSR reactor by spectral distri­ experiments and optical simulations [38]. Ma et al. compared the
bution method [31]. Reflective concentrators such as parabolic trough average collection efficiency and stable operation time of a cylindrical
concentrators (PTC), solar dish concentrator, compound parabolic Fresnel lens-based system coupled to a vapor generator with a parabolic
concentrators (CPC), and linear Fresnel reflectors (LFR) are technically trough type concentrator [39]. In another paper, the geometrical pa­
mature and conducive to scale-up. MSR systems based on slot concen­ rameters of the cylindrical Fresnel lens were optimized through theo­
trators face problems such as local overheating caused by uneven input retical analysis, experimental tests, and indirect estimation [40].
solar flux [32]. Maintaining a high methanol conversion efficiency re­ However, the work combining the cylindrical Fresnel lens with the
quires longer concentrator and reactor diameter lengths due to the solar-thermal-chemical process has not been reported, even though it
limitation of their concentration ratios. In short, more module installa­ shows remarkable potential.
tion area is required [33]. In addition, solar power states located in East The literature reviewed above has presented various system con­
Asia and some Central Asian regions, although rich in solar resources, struction methods and optimization ideas for solar-driven thermo­
also face the dilemma of wind and sand contamination on the surface of chemical reactions to hydrogen production. In this paper, a novel MSR-
solar concentrators [34], and large area installation of PTCs will un­ STR system is proposed to study the comprehensive reaction charac­
doubtedly increase labor and economic costs. teristics and system performance, and the main work is organized as (1)
Another idea is to couple solar-driven thermochemical hydrogen A comprehensive three-dimensional model is established that in­
production components with other novel concentrating collector sys­ corporates the geometrical, optical, thermal, and chemical characteris­
tems. Several innovative concentrator designs based on Fresnel lenses or tics. (2) The performance analysis of this MSR-STR system includes an
reflectors have been proposed to miniaturize the modules and maintain optical model based on the Monte Carlo ray tracing (MCRT) method, a
efficiency while facilitating installation and post-maintenance [35]. validated chemical reaction kinetic model based on the computational
Compared to Fresnel reflectors, Fresnel lenses can obtain high geometric fluid dynamics (CFD) model, as well as the economic analysis and
concentration ratios and thus high-temperature heat. Zhang et al. evaluation indicators. (3) Numerical analysis of the solar-driven MSR-
designed a cylindrical photothermal reactor that combined STR system was carried out to obtain better optical efficiency and

2
N. Zhao et al. Renewable Energy 222 (2024) 119952

investigate the H2 production performance. convection and the MSR series reaction, (3) the heat from the outer wall
of the glass is transferred to the inner wall of the glass hood by radiative
2. System and methodology description heat transfer. The heat is then transferred to the metal tube walls by
surface-to-surface (S2S) radiative heat transfer within the near-vacuum
2.1. Physical model of MSR-STR system environment.
However, the overall MSR reactions are complicated to fully consider
Fig. 1 illustrates the schematic diagram of the MSR-STR reaction all aspects of the methanol-steam reforming process. Therefore, some
system. The optical components consist of a concentrator, parabolic assumptions have been made in this study. The S2S model is chosen as
reflector, solar thermal reactor, and supporting structure, utilize a the main thermal transfer path, and the annulus space is considered as a
single-axis solar tracking system driven by a stepper motor to track the perfect vacuum to ignore the effect of convective heat transfer. The gas
sun. The MSR-STR system concentrates solar rays using an LCFL, as mixture, including reactants and products, is assumed to be an ideal gas.
depicted in Fig. 1(a). The incident solar rays are subsequently irradiated The catalyst particles are assumed to be homogeneous and uniformly
onto the STR surface and absorbed by selective absorbing coatings. The filled throughout the reactor. Optical simulations are conducted using
CPC is used to collect solar rays and homogenize the solar flux distri­ the MCRT method with 105 randomly emitted rays. The effect of solar
bution on the reactor surface. The STR is designed with a tubular length azimuth on the convergence process is also neglected in the subsequent
(L) and an outer glass tube diameter (dgo), and the thickness of the glass optical simulation.
tube and the diameters of the outer and inner steel tube (dso and dsi), as
presented in Fig. 1(b). 2.3. Mechanism of MCRT method
The annular solar thermal reactor, as depicted in Fig. 2, is the
essential component of the endothermic MSR reactions. The 3-D STR The MCRT method is used to analyze the designed path of the solar
model mainly consists of three computational parts: (1) the stainless rays and the effective optical efficiency of the MSR-STR system. The
tube and the glass cover of the reactor, (2) the porous bed domain of the concentrated solar flux is calculated by the MCRT optical model based
Cu/ZnO/Al2O3 catalyst and muti-component gas mixture, and (3) the on the Law of Large Numbers [41]. The ray tracing is based on the
vacuum annular space between the receiver tube and the glass tube. Tracepro software, and the material, light source, and operation prop­
Three main reactions were adopted as the main scope of MSR reaction erties has carefully reviewed. The PMMA is obtained as the material of
kinetics that is generally accepted for methanol decomposition. The the linear cylindrical Fresnel concentrator (relative refraction index =
reactions inside the STR involve a series of solar-thermal-catalytic sur­ 1.5), and the spectrum transmittance of PMMA is given in Table 1. The
face reactions, and the design of the concentrator configuration, optical fluid reactants entered the reactor is heated by the irradiance temper­
calculations, photothermal conversion, thermal mass transfer, and ac­ ature. Thus, the concentrated solar radiance is not only considered as the
curate simulation of the reaction kinetics are all considered in this study. boundary condition in the radial direction of the reactor, but also as a
heat source in the porous catalyst filled bed to promote the reaction
2.2. Thermal and mass transfer process of MSR-STR system forward.

The solar thermal reactor mainly exchanges heat and mass with the 2.4. Numerical methods
surrounding environment through various processes: (1) a portion of the
light beam that passes through the cylindrical Fresnel concentrator is 2.4.1. Governing equations
directly irradiated onto the upper surface of the reactor, while the rest of The realizable k-ε two-equation turbulence model has been accepted
the light is reflected by the secondary concentrator mirror and then for most common engineering solves [42]. The continuity equation,
irradiated onto the side and lower surfaces of the reactor. The absorbed momentum conservation equation and energy conservation equation of
coating on the reactor wall converts the light into heat energy, which the realizable k-ε turbulence model are listed below. The porous catalyst
heats the reactor, (2) the heat energy from the reactor wall is transferred filled bed inside the STR consists of the solid phase (i.e., the Cu/Z­
to the interior of the reactor and the porous Cu/ZnO/Al2O3 catalyst bed nO/Al2O3 catalyst particles) and the gas phase mixed with MSR re­
by heat conduction, followed by heat transfer between the porous cat­ actants and/or products (i.e., CO, CO2, H2, CH3OH).
alytic particles and the multi-component gas mixture by thermal k equation:

Fig. 1. The Schematic diagram of a cylindrical solar-thermal-reaction system.

3
N. Zhao et al. Renewable Energy 222 (2024) 119952

Fig. 2. Structure of an annular solar methanol-steam reforming reactor.

Energy conservation equation (porous phase):

Table 1
( ) ( )
Spectrum transmittance of the PMMA. ∇ ρcp →
u T = ∇ λeff ∇T + St (9)
Wavelength λ/nm Transmittance/% Wavelength λ/nm Transmittance/%
Energy conservation equation (fluid phase):
360 5 900 90.86 ( ) ( )
370 22.4 1000 91.27 ∇ ρcp →
u T = ∇ λf ∇T (10)
380 59.6 1100 91.57
390 83.49 1200 83.98 where the heat capacity cp and the thermal conductivity λ of the mixture
400 89.81 1300 91.25
can be expressed by Eq. (11) and Eq. (12), respectively. The thermal
500 91.82 1400 72.84
600 92.07 1500 87.6 conductivity of filled bed is given as Eq. (13), and the source term of the
700 92.18 1600 83.53 chemical reactions is shown as Eq. (14) [44].
800 92.36 1700 24.19 ∑
cp = mi cp,i (11)
i
[( ) ]
μ
∇(uk) = ∇ μ + t ∇(k) + Gk − ρε (1) ∑ Mi λi
σk λ= ∑ (12)
i
Mj φij
ε equation: j

[( ) ] ( )
∇(ρuε) = ∇ μ+
μt
∇ε + c1 ρSε − c2 ρ
ε2
√̅̅̅̅̅ (2) λeff = εp λf + 1 − εp λ (13)
σε k + νε
( ∫T )
∑ ho,i
Continuity equation: St = − +
(
cp,i dT Ri,r
)
(14)
MW,i Tref,i
∇ ⋅ (ρ→ (3)
i
u )=0
Momentum conservation equation (for chemical species): where ho,i is the enthalpy of species i, Ri,r is the volumetric mass rate of
[ ] creation of species i, which can be automatically calculated by the
( ) 2
ρ(→
→
u = ∇ ⋅ − p I + μ ∇→
u ⋅ ∇)→ u + (∇→
u )T − μ(∇ ⋅ →
→
u ) I + ρgi + Fi ANSYS FLUENT software [42]. Here, we defined i = 1, 2, 3 for the MSR’s
3 sub-reactions.
(4) The component transport equations should also be solved for this
endothermic MSR reactions in STR. The local mass fraction of the
where the source term Fi is valid only for the homogeneous porous reactant/product species is calculated in the FLUENT software with a
catalyst. Fi can be expressed as follows: convection-diffusion equation, which can be shown as Eq. (15) [42]:
( )
μ 1
Fi = − ui + c2F ρ ± |u|ui (5) ∂
(ρui mn ) = −
∂Jn
+ Sn,R (15)
c1F 2 ∂xi ∂xi
where the c1F and c2F locate in the viscous loss term and inertial loss term where the Jn is the mass diffusion flux of species n, Sn,R is the source term
are the permeability and the inertial resistance factor [42]. These of the net rate of production species n. The Jn in a turbulence flow can be
experimental based terms are given as follows: expressed as Eq. (16):
D2p ε3p ( )
(6) μ ∂mn
c1F = ( )2 Jn = − ρDn,m + t (16)
150 1 − εp Sct ∂xi
( )
3.5 1 − εp where Dn,m is the mass diffusion coefficient of species n, Sct is the
c2F = (7) effective turbulent Schmidt number, commonly set as 0.7 [42]. In
Dp ε3p
addition, the pore diameter varies with the radial position in the actual
where εp is the catalyst bed porosity. Dp is the average diameter of the non-uniform distribution conditions [45]. The Maxwell-Stefan model is
Cu/ZnO/Al2O3 catalyst particles [43]: used to derive relations for the diffusion fluxes. It should be noted that
the porous catalyst has simplified as a pseudo homogeneous medium.
εp = 0.375 +
0.34
/ (8) More accurate diffusion modeling for the MSR response will be inves­
Di Dp tigated in future work.
Considering the porous characteristics of the Cu/ZnO/Al2O3 catalyst The Sn,R can be calculated as follows:
bed and the chemical species of the mixture gas, the energy conservation ∑
Nr

of the porous phase and fluid phase can be expressed as: Sn,R = Mw,n Rn,r (17)
r=1

4
N. Zhao et al. Renewable Energy 222 (2024) 119952

where Mw,n represents the molecular weight of species n, and the Rn,r is listed in Table 2. It should be pointed out that the thermophysical pa­
the molar rate of each species in the specified reaction r, which are rameters, includes heat capacity and thermal conductivity of pure gas
demonstrated in Section 2.4.2. are varies with the working temperature. A common treatment is to
express the physical parameters of the substance in the form of poly­
2.4.2. Comprehensive kinetic model of MSR reaction nomial functions, expressed as Eq. (24) and Eq. (25) [47].
The methanol-steam reforming using Cu/ZnO/Al2O3 catalyst typi­ cp,i = Ai,0 + Ai,1 Tf + Ai,2 Tf2 + Ai,3 Tf3 + Ai,4 Tf4 (24)
cally consists of three sub-reactions, methanol steam reforming reaction
(MSRR), methanol decomposition reaction (MDR), followed by the λi = Bi,0 + Bi,1 Tf + Bi,2 Tf2 + Bi,3 Tf3 (25)
reverse water-gas shift reaction (rWGSR). The overall reactions are
expressed as Eqs. (18)–(20): The Ai,0− 4 and Bi,0− 3 are the heat capacity coefficient and the thermal
Methanol steam reforming reaction (MSRR): conductivity coefficient of species i, where i = 0, 1, 2, 3, 4 is the species
of CH3OH, H2O, H2, CO, and CO2, respectively. The details of the heat
CH3 OH + H2 O⇌CO2 + 3H2 ΔHo298 = +49.4 kJmol− 1
(18)
capacity coefficient and the thermal conductivity coefficient are listed in
Methanol decomposition reaction (MDR): Table 3.
The mixing law is adopted in to calculate the heat capacity of mixture
CH3 OH⇄CO + 2H2 ΔHo298 = +90.64 kJmol− 1
(19)
gas, and the mass weighted law is used to calculate the thermal con­
Reverse water-gas shift reaction (rWGSR): ductivity of mixture gas, respectively:

(20) ∑ ∑
1 n n
CO + H2 O(g)⇄CO2 + H2 ΔHo298 = − 41.1 kJmol−
cp,m = yi cp,i ; λm = yi λi (26)
Peppley et al. [46] have proposed a Langmuir-Hinshelwood (LH) i i=1

model based on the dehydrogenation of adsorbed methoxy as a

rate-determining step, shown in Eq. (21) - Eq. (23). For the reaction rate 2.4.3. Boundary conditions
calculation of MSR reaction on the catalyst-packed bed, this widely used The mean boundary conditions are set in Table 4. The S2S thermal
LH model is adopted to describe the kinetic characteristics of the re­ radiation was implemented in FLUENT for near vacuum annulus space
actions comprehensively. between the steel tube and the glass cover. The emissivity of the glass
( ) cover is assumed as a constant value of 0.9, while the emissivity of the
pCH3 OH p3H pCO2 STR absorbing coating surface is set as a near-linear user defined func­
kR KCH3 O (1) 1− CST1 CST1a
tion (UDF) of εao = 4.2e− 4 Tao(K) − 0.0995 [49], and compiled with the
2
1/2 KR pCH3 OH pH2 O
pH
2
rMSRR = ( )
( ) macro DEFINE_PROFILE in C. All end surfaces are defined as adiabatic
1 + KCH3 O(1)
pCH3 OH
1/2 + KHCOO(1) pCO2 pH2 + KOH(1)
1/2 pH2 O
1/2
1/2 1/2
1 + KH(1α) pH2 wall boundaries because of the metal-glass joints and insulation fillers at
both ends of the STR. In addition, species mass fractions at the fluid inlet
pH pH
2 2

(21) surface of the reactor are calculated with the molecular weights of steam
( ) to methanol and their corresponding molar ratio.
pCH3 OH p2H pCO
kD KCH3 O(2) 1− 2
CST2 CST2a
2.4.4. Solution conditions
1/2 KD pCH3 OH
pH
rMDR = ( 2
) (22)
pCH3 OH
(
1/2 1/2
) The governing equations along with the boundary were solved
1 + KCH3 O(2) 1/2 + KOH(2) pH2O
1/2 1 + KH(2α) pH2 simultaneously based on the FLUENT 2020R2 [42]. The solver settings
pH pH
in the FLUENT is shown in Table 5. The solver governing equations of
2 2

pCO pH2 O
(
pH2 pCO2
) the three-dimensional STR model were discretized by the finite volume
kW KOH(1) 1/2
pH
1− KW pCO pH2 O
CST1 CST1 method. The convection terms of governing equations are discretized by
rrWGSR = ( 2
(23) the second-order upwind scheme. The PRESTO! (PREssure STaggering
Option) scheme is used for pressure gradient terms solving. Due to the
pCH3 OH 2
1/2
1 + KCH3 O(1) 1/2 + KHCOO(1) pCO2 pH2 + KOH(1) pH2O)
1/2
precision and solution time consideration, the iterative solution is suf­
pH pH
2 2

ficiently convergent until the convergence criterion is satisfied. The

where rMSRR rMDR, and rrWGSR are volumetric mole reaction rates of three normalized residuals of continuity are below 10− 5, the residuals of the
sub-reactions, pi is the partial pressure of species i. Other parameters are other equations are below 10− 7.
Convective heat transfer and heat conduction between the multi-

Table 2
Parameters for reaction kinetic model of methanol steam reforming [46,48]. Table 3
The coefficient of heat capacity and thermal conductivity for gas species.
Parameters Value Parameters Value
Coefficients CH3 OH H2O H2 CO CO2
kR/m2 s− 1 7.4 × 1014e102,800/ KH (1) e− 100.8/R+50,000/
\Species
mol− 1 (RT) (RT)

kD/m2 s− 1 3.8 × 1020e170,000/ KH (2) e− 46.2/R+50,000/

Ai.0 5.0805E5 1563.077 13602.45 968.390 429.929
mol− 1 (RT) (RT)
Ai.1 − 5542.193 1.604 3.402 0.449 1.874
kW/m2 s− 1 5.9 × 10 e 13 87,600/(RT)
KR 10 − 3066/T+10.592
Ai.2 22.5950 − 2.933E- − 3.358E- − 1.152E- − 1.949E-
mol− 1 3 3 3 3
− 41.8 R− 20,000/(RT)
KCH3O (1) e / KD 10− 5139/T+12.621 Ai.3 − 0.04064 3.216E-6 − 3.908E- − 1.657E- 1.297E-6
KCH3O (2) e30/R− 20,000/(RT) KW 102073/T− 2.029 7 6
KHCOO e179.2/R− 100,000/(RT) Cs1, Cs2/mol m− 2 7.5 × 10− 6 Ai.4 2.7267E-5 − 1.157E- 1.705E-9 − 7.346E- − 3.999E-
KOH (1) e− 44.5/R+20,000/(RT) Cs1a, Cs2a/mol 1.5 × 10− 5 9 10 10
m− 2 Bi.0 0.1856 7.341E-3 8.099E-3 5.067E-4 − 7.215E-
KOH (2) e30/R+20,000/(RT) h 0,CH3OH /kJ mol − 220.94 3
− 1
Bi.1 5.4033E-4 1.013E-5 6.689E-4 9.125E-5 8.015E-5
− 1
h 0,H2 O /kJ mol − 241.81 h 0,H2 /kJ mol 0 Bi.2 − 2.32E-6 1.801E-7 − 4.158E- − 3.524E- 5.477E-9
− 1
7 8
− 1
h 0,CO2 /kJ mol − 393.51 h 0,CO /kJ mol − 110.53 Bi.3 2.2667 9.100E- 1.562E- 8.199E- − 1.053E-
− 1
E− 9 11 10 12 11

5
N. Zhao et al. Renewable Energy 222 (2024) 119952

Table 4
ηo = (Qrec /Qtot ) × 100%
Boundary conditions of the FVM method based MSR comprehensive model. (29)
ηt = (Quse /Qrec ) × 100% = Quse /(Quse + Qloss ) × 100%
Boundaries Type Expressions
Methanol conversion rate (ηc), the H2 yield (YH2 ) and the CO selec­
Inlet Constant mass flow and qx = qm,in, qy = qz = 0; T = Tin
temperature tivity (SCO) are used to assess the overall MSR performance. Here, the
Outlet Constant outlet pressure p = pout = 1.01 bar methanol conversion rate is defined as the ratio between converted
Reactor wall (end Coupled no-slip for The no-slip shear condition, zero methanol and the inlet methanol flow rate (nCH3OH, in) [51]:
surface) velocity and adiabatic for heat flux
heat transfer nCH3 OH,in − nCH3 OH,out
ηc = (30)
Glass cover (radial Coupled no-slip for The no-slip shear condition, nCH3 OH,in
surface) velocity and solar flux for concentrated solar flux
heat transfer distribution by MCRT model YH2 is defined as the ratio of the H2 production rate to the inlet methanol
End surface of Coupled no-slip for The no-slip shear condition, zero
flow rate:
glass cover and velocity and adiabatic for heat flux
reactor heat transfer nH2
YH2 = (31)
nCH3 OH,in

SCO is the ratio of the molar fraction of produced CO to that of all

Table 5
produced carbon-containing species of both CO and CO2 in MSR system.
Solver settings.
nCO
Flow solver Segregated SCO = (32)
nCO + nCO2
Pressure-velocity coupling PISO (Pressure-Implicit with Splitting of Operators)
scheme
where nCO and nCO2 are molar flow rate of produced CO and CO2,
Discretization scheme 2nd order upwind (for convection terms);
PRESTO! (Pressure staggering option) scheme (for respectively.
pressure gradient terms)
3. Results and discussion

component gas mixture in the porous medium and the catalyst surface 3.1. Model validation
become the dominant modes of heat exchange, so that heat radiation
between catalyst particles is neglected. To simplify the analysis, several The feasibility and accuracy of the proposed kinetic reaction model
assumptions are made in the numerical simulation: (1) The gas mixture and numerical method were confirmed by comparing the simulation
is assumed to be in an ideal gas state and flow in the STR is considered to results and the experimental results. Peppley et al. [52] has conducted a
be in a steady state; (2) The catalyst is assumed to be isotropic and set of experiments of MSR reactions in a cylinder reactor packed of
porous; (3) Radiation heat transfer in the catalyst is neglected; (4) The Cu/ZnO/Al2O3 catalyst particles. The reactor’s nominal inner diameter
gas and solid phases within the catalyst are assumed to be in local (I.D.) is 2.21 cm, and Ref has given other typical baseline reaction
thermal equilibrium and the temperature differences are ignored. conditions [52]. To be close to the experimental conditions of the real
reaction, the uniform wall temperature (513 K and 533 K) setting similar
to the experiment is adopted in the reactor circumference. It should be
2.5. Parameter valuation noted that the central temperature in the experiment is obtained by the
thermocouple placed in the center of the reactor, rather than the tem­
Some significant parameters have been proposed to comprehensively perature of the reactor wall. Fig. 3 illustrates the comparing results of
evaluate the performance of MSR hydrogen production system. The the values of methanol conversion rates ηc. Here, the weight flow ratio
optical and thermal characteristics, fluid dynamics, and component re­ (WFR) denotes the ratio between the mass weight of the catalyst parti­
action performance of the MSR-STR are compared and analyzed. cles and the methanol molar flow rate at the inlet. The curves of
The Reynolds number, average Nusselt number and average friction
factor are the important evaluate factor for heat transfer and fluid dy­
namics, which are given as [50]:
ρuDsi hDsi ΔpDsi
Re = , Nu = ,fm = (27)
μ λ (1/2)ρu2

where h is the average heat transfer coefficient (h = A×(TQ1u− T0 )), Δp is the

pressure difference between reactor inlet and outlet. The expression of
thermal conductivity λ is presented as Eq. (25) shows.
The optical and thermal performance of the MSR-STR system in­
cludes the average concentrating solar flux density (qa,m), nonunifor­
mity flux distribution (σq), and optical and thermal efficiency, shown as
Eq. (28).
∑
Nc Nc ⃒
∑ ⃒
qi ⃒ qi − qa,m ⃒
qa,m = i=1 , σq = i=1 × 100% (28)
Nc Nc ⋅qa,m

where Nc is the grid number at the methanol reforming reactor surface,

and qi is the local energy flux at the ith grid.
The optical efficiency (ηo) of the STR is specified as the ratio of the
received solar energy (Qrec) to the total incident solar energy (Qtot), and
the thermal efficiency (ηt) is defined as the ratio of the useful solar en­ Fig. 3. Comparison of methanol conversion rate numerically predicted in this
ergy (Quse) to the received solar energy (Qrec). study with the experimental data measured by Peppley et al. [52].

6
N. Zhao et al. Renewable Energy 222 (2024) 119952

numerical methanol conversion rates show a good agreement with the Table 7
experimental data for both reaction conditions at 513 K and 533 K. The Grid independence checking for a basic computational case with evaluation
maximum deviation between the numerical and experimental data is parameters.
10.7 %, which occurs when WFR is 3.2 and the reaction temperature is Cell number Tf,max /K Nu YH2 SCO fm ηc
513 K. The average deviation is 7.8 %. Considering the differences be­
3845522 513 8.71 0.013 0.0012 17.276 0.019
tween numerical prediction and experimental operation, this deviation 4034069 513 8.72 0.013 0.0013 17.475 0.022
is acceptable to demonstrate the feasibility and reliability of this kinetic 4556124 513 8.72 0.015 0.0013 17.477 0.022
reaction model. 5395986 513 8.73 0.014 0.0013 17.475 0.021
The validated kinetic model was embedded in STR to analyze the 6023453 513 8.72 0.015 0.0013 17.483 0.022

methanol-steam reforming reactions. Here, the STR coupled with a cy­

lindrical Fresnel lens was used to analyze and optimize the character­
istics and performance of the novel MSR hydrogen generation system.
The elements such as system structural optimization, inlet methanol
molar rate, steam-to-methanol ratio (SMR), and operation inlet tem­
perature were taken into consideration. The detailed numerical pa­
rameters of the chemical reaction kinetic model are presented in Table 6
with different STR deposition locations, WFR, and other basic compu­
tational cases. In addition, the catalyst particle diameter is set as 4.853
× 10− 3 to ensure the porosity while avoid the diffusion gradient of the
catalyst filled bed. The particle size is a crucial parameter of species
diffusion. The bigger the particle size is, the great the gradient of pres­
sure, temperature and diffusion coefficient [45,53].
The structured mesh was established for the STR, and the mesh
quality was verified through grid independence analysis. The grid in­
dependence verification is calculated based on the boundary conditions
listed in Table 7. Five mesh models were used for grid independence
checks with the basic computational case, and the subtle characteristics
in the near-wall region were compared to balance the computational
resources with the reliability of results. It is worthy to pointed out that
the recommended grid growth ratio of the neighboring grid should be
greater than 1.1 [54]. The evaluation results of the parameters indicate
that the grid system with 4,556,124 cells is sufficient for the analysis of
Fig. 4. Concentrated energy flux distributions of the circumferential
the solar thermal reactor. After confirming the grid quality, the nu­
STR surface.
merical cases were processed, and the evaluation parameters were used
to illustrate the cases and discuss the results.
between the center of the STR cross-section and the center of the sec­
ondary reflector cross-section was increased from 0 to 50 mm to enhance
3.2. Modeling of the concentrating flux nonuniformity the uniformity of the energy flux density on the STR surface. It should be
noted that changing the imaging distance also alters the width and
3.2.1. Effect of concentrated solar flux distribution inclination angle of the mirrors on both sides of the linear Fresnel lens.
Fig. 4 presents the energy flux distribution on the circumferential The initial structure of Case 1 exhibits a significant non-uniformity of
surface after concentration. Here we compare the surface energy flux energy flux. This non-uniformity creates a potential thermal sintering
distribution of the reactor corresponding to four optical assemblies: Case region for the catalyst particles at the center of the upper semicircular
1 presents a flux distribution based on a single linear Fresnel concen­ circumference of the STR. Case 2 presents a certain degree of improve­
trating structure; Case 2, a simple linear Fresnel concentrating structure ment in flux uniformity. However, the optimized Case 3 yields the most
coupled with CPC; Case 3, a concentrating structure with both a reflector uniform flux distribution over 180◦ of the STR’s circumferential surface
and a concentrator. Three reflect mirrors were added on each side of the by maintaining flux density above a reference value. Increasing the
original Fresnel lens to increase the concentrating area while reducing receiving half-angle of the secondary concentrator not only collects
the optical loss. The primary concentrator module’s width was increased more incident light but also reduces the overall aspect ratio of the
from 1300 mm to almost 2000 mm, resulting in a geometric concen­ concentrator. With a uniform distribution, the flux density around the
tration ratio of about 33 and a maximum optical efficiency exceeding 91 STR surface is close to 9 kW/m2, which satisfies the photothermal
%. conversion (Case 1) and drives the thermochemical reaction forward
Case 3 optimizes the imaging distance based on Case 2 and increases [50]. In a subsequent study, we adopted the structure of Case 3 to
the CPC receiving half-angle from 45◦ to 60◦ . Moreover, the distance improve the reaction rate and uniformity while ensuring the safe
conduct of the reaction.
Table 6
Parameters of the MSR-STR system for numerical analyze. 3.2.2. Effect of longitudinal incidence angle
Parameters Values Parameters Values The uniformity of the spot on the receiving surface is influenced by
the longitudinal incidence angle (β) of the linear Fresnel concentrator.
W/m 1.3 L/m 8.0
f/m 1.92 Dc/m 2.0
Fig. 5(a) illustrates the concentrator’s focal point variation at different
ρp/kg m− 3 1300 Dp/m 4.853 × 10− 3 incidence angles. As the θ increases, the refracted light rays converge
Dgo/m 0.115 Dso/m 0.07 earlier below the concentrator, and the focal point moves away from its
Dgi/m 0.110 Dsi/m 0.066 designated position, causing the light beam to gradually diverge and
Dm/m2s− 1 6.8 × 10− 5
Tin/K 473
become disordered. After simulating the system’s longitudinal incidence
pout/MPa 0.1 a/m 1.3
n 1.59 δg 0.0025 angle, the results are depicted in Fig. 5(b). When the incidence angle θ is
δl,max/m 0.007 δl,min/m 0.002 less than or equal to 20◦ , most of the incident light can still reach the

7
N. Zhao et al. Renewable Energy 222 (2024) 119952

Fig. 5. The longitudinal incidence angle corresponds to the change of light path and focal length. (a) Light path and focusing points. (b) Focus characteristics changes
with longitudinal incidence angle.

absorber due to the dual effect of the secondary reflector and the opti­ receiving width and spills out from the edge of the secondary concen­
mized receiving distance. However, when the θ exceeds 20◦ , the focus trator. When the γ increases to 1.5◦ , there is a significant decrease in the
length shifts significantly upward, causing the focus width to increase optical efficiency of the system.
sharply, while the light divergence becomes serious. Considering the
installation and operational characteristics of the equipment, it is pref­ 3.3. Operating parameter characteristics analyze
erable to install the MSR-STR system at low latitudes to avoid excessive
solar axial incidence inclination. 3.3.1. Reactor wall temperature
Fig. 7 displays the comprehensive performance of the MSR-STR
3.2.3. Tracking precision requirement system under various computational conditions as listed in Table 4.
Axial tracking accuracy plays a crucial role in the optical efficiency The wall temperature of the STR varies automatically with the direct
and energy flux distribution uniformity of the MSR-STR system. Fig. 6 normal irradiance (DNI) per hour at Hohhot, China (41 ◦ N, 110 ◦ E) from
shows the system optical efficiency when the imaging distance is 1920 9:00 a.m. to 3:00 p.m. The maximum temperature (Tf, max) in the filled
mm, and the longitudinal offset is 0◦ . The imaging focus gradually shifts bed was monitored to prevent catalyst thermal sintering. The thermal
away from the STR cross-section’s center, and the system’s optical ef­ variation in the reaction zone is caused mainly by the inlet flow rate
ficiency decreases as the tracking error angle (γ) increases from 0◦ to (WFR), the wall heat flux density of the STR, the radiation heat dissi­
2.5◦ . With an increasing γ, more rays initially reach the inner surface of pation from the STR to the atmosphere, and the reaction heat absorption
the secondary concentrator, leading to direct reflection and subsequent inside the STR. When the irradiation intensity exceeds the pre-set value,
concentration on the STR, causing a decline in the light intensity. Case 3 requires more attention to Tf, max than Case 0. From Fig. 7(a) it
Additionally, as γ increases, more light deviates from the designed can be seen that as the WFR value decreases, the inlet mass flow rate
increases and the maximum temperature Tf, max inside the STR filled bed
decreases while the DNI reaches to794 W/m2. Variations of Nu with
WFR shows that Nu decreases with the increasing of WFR, and fm value
shows the opposite trend with Nu. The smaller Nu presents the gentler
the flow in the STR. Although it slows down the intensity of convective
heat transfer to some extent, it also means that the reactants stay in the
STR for a longer time. The maximum mass flow rate allowed in the
reactor at this point is 16.42 g/s. The corresponding methanol conver­
sion and H2 yield within the STR are shown in Fig. 7 (b). It is apparent
that the methanol conversion on this typical day at the time of maximum
solar irradiation cannot reach 100 % due to the constraint of the
maximum temperature.
Fig. 8 illustrates the response characteristics of the MSR-STR system
on a typical day on an hourly basis. The optical efficiency of the system
exhibits a pattern of low values on the two ends and high values in the
middle. The optical efficiency is calculated based on the solar altitude
angle, while the thermal efficiency considers the radiation heat loss of
the STR and atmospheric ambient. It is worth noting that the current
calculation conditions do not consider the influence of the azimuth angle
on the system efficiency. Although the WFR is increased around 12:00 a.
m. to prevent catalyst particle sintering, the axial incidence angle of the
sun remains the primary factor limiting the reaction rate.

Fig. 6. Optical efficiency with various track error angle.

8
N. Zhao et al. Renewable Energy 222 (2024) 119952

Fig. 7. Comprehensive characteristics of the MSR-STR system with varying WFR. (a) WFR with Nu, fm and permissible temperature. (b) WFR with YH2 , SCO and
methanol conversion rate.

Fig. 8. Performance of MSR-STR system with varying irradiance. (a) Variations of irradiance, irradiated energy, and efficiency (b) Variations of ηc, H2 molar flow
rate and WFR.

3.3.2. Inlet mixture gas temperature 3.3.3. Reactant feed ratio

The MSR-STR hydrogen production system boasts favorable geom­ The SMR has a significant effect on the hydrogen yield of the MSR-
etry and reaction conditions but requires reactant heating and gasifi­ STR reaction system. Fig. 10 illustrates the effect of SMR on the con­
cation before flowing through the porous catalyst. The temperature of version capacity of CH3OH and the production of H2. As the SMR in­
the gasified mixed methanol-water vapor entering the reactor has a creases from the default of 1.0–2.0, the conversion of CH3OH increases
significant effect on the thermochemical reaction. To predict the optimal accordingly, but the molar fraction of H2 shows a gradual decreasing
hydrogen production rate, we established a range of operating condi­ trend. Based on the MSR reaction model described in subsection 2.4.2,
tions. Fig. 9 presents a histogram illustrating the impact of different gas the production rates of each species and reaction energy in the MSR
mixture temperatures and methanol molar flow rates at the STR inlet on process are organized in Table 8. According to the equations of CH3OH
the H2 production rate. Higher reaction inlet temperature (513 K) and consumption rate and H2 production rate, more water vapor as the SMR
larger WFR (110 kg s/mol) correspond to higher hydrogen yields (0.77 gradually increases will result in a larger share in the mixture of prod­
mol/s) for the MSR-STR system. Neither the mixture gas temperature ucts and reactants, which will reverse the dilution of the H2 concen­
nor the methanol molar flow rate in inlet correlates linearly with the tration in the mixture.
hydrogen molar production rate. This is because as the STR reaction
proceeds, the inlet gas temperature gradually rises due to the glass tube 3.4. Economic analysis
reaching thermal equilibrium with the outside world. A lower inlet gas
temperature retards this temperature rise for a given molar flow rate, The cost analysis of an MSR-STR system was performed that employs
and a larger inlet molar flow rate has the same temperature rise sup­
a cylindrical Fresnel lens heat collection area of 16 m2. Table 9 shows
pression effect. Additionally, larger WFR and gas temperatures tend to the detailed operating conditions and estimated costs of the system. The
correspond to higher hydrogen production rates in the given range,
cost mainly consists three parts: total capital investment (CTC), annual
consistent with the mechanism of MSR reaction kinetics. operation and maintenance cost (CO&M) and the reactant cost (CMethanol).
For the MSR-STR hydrogen production system developed in this study,

9
N. Zhao et al. Renewable Energy 222 (2024) 119952

Table 9
Estimated levelized cost of hydrogen of the system.
Parameter Value

System size, energy, and materials:

Light receiving area 16 m2
Available solar irradiance energy 35.9 MWh/yr
Methanol requirement 90.4 tons/yr
H2 yield 57.3 tons/yr
Capital cost:
Solar thermal reactor $ 22/m2
Linear Fresnel concentrator $ 480/m2
Other accessories $ 500
Primary capital cost $ 8532
Annual investment cost:
Annual Fixed Assets Rate 5%
Operation & maintenance $ 1440/yr
Methanol cost $ 33448/yr
Total annual cost $ 34888/yr
Levelized Costs:
LCOH $ 1.77/kg

the levelized cost of hydrogen (LCOH) is calculated as [26]:

fCR ⋅CTC + CO&M + CMethanol
LCOH = (33)
Eannual

where CTC is the total investment cost, CO&M and CMethanol are the
Fig. 9. H2 yield change with mixture gas temperature and CH3OH inlet annual O&M cost and methanol cost, respectively. Eannual is the annual
flow rate. H2 production (kg). The capital recovery factor fCR can be calculated by
the following equation:
[ ]
(1 + i)n − 1
fCR = (34)
i⋅(1 + i)n

where i is the annual fixed interest rate and n is the payback period.
The cost of the cylindrical Fresnel concentrator is approximately
$7680 [39]. According to a report by NREL, the unit footprint cost of a
vacuum heat collector used as a STR is roughly $22/m2 [56]. Thus, the
initial investment cost of the MSR-STR system is estimated to be $8532.
The annual fixed interest rate is assumed as 5 % and an O&M rate is
taken as 2 % [26]. Maintenance costs are the sum of routine mainte­
nance costs and repair and replacement costs. The annual O&M cost also
includes the annual catalyst replacement cost, which is $38.7/kg for
CuO/ZuO/Al2O3. According to the above costs analysis, the O&M cost is
$1440 per year. The annual DNI of Hohhot city with 3560 effective
hours is depicted hourly in Fig. 11. The annual DNI data was measured
and collected with a weather station located on Hohhot, China. As a
result, it costs $33,448 per year to utilize roughly 90.4 tons of methanol.
In addition, the small structure allows it to be installed on a building’s
roof or an open space for direct use, without considering the trans­
portation cost of methanol and hydrogen. Furthermore, the carbon
reduction gain from using solar energy and the gas separation and car­
Fig. 10. Outlet mole fraction of CH3OH and H2 under various steam methanol bon capture costs of the STR export syngas is not considered.
ratio. MSR = 55 kg s/mol, Tw = 513 K. Assuming a payback period of 10 years, the LCOH of the system is
estimated to be $1.77/kg after considering the cost of all reactants and
materials required for the system, which is competitive with industrial
Table 8 MSR reactions under similar annual solar irradiation conditions.
Reaction rate and reaction heat in STR [55]. Furthermore, since the methanol is not fully reacted in the STR all the
Generation rates Expression time, the methanol exported from the outlet mixture gas can also be
CH3OH consumption rate rCH3 OH = − rR − rD reused as feedstock, which will further reduce the cost of material usage.
H2O consumption rate rH2 O = − rR − rWGS
CO2 production rate rCO2 = rR + rWGS 3.5. Discussion, foresight and outlook
CO production rate rCO = rD − rWGS
H2 production rate rH2 = 3rR + 2rD + rWGS
Energy source of MSR reactions qc = − rR ΔHR − rD ΔHD + rWGS ΔHWGS Overall, this study proposes a novel green route for solar-driven
thermochemical hydrogen production and emphasizes the feasibility
The ΔHR , ΔHD and ΔHWGS are the heats of reaction for the reforming, decom­
of further research and development in this field. However, more studies
position and WGS reactions, respectively.
could be carried out to combine the real operation conditions, includes
real gas model, time-dependent transient flow dynamics, particle-filled
catalyst bed instead of porous medium, etc. Further verification and

10
N. Zhao et al. Renewable Energy 222 (2024) 119952

Fig. 11. Annual direct normal radiation intensity in Hohhot.

additional experimental studies are essential to thoroughly verify the found yet. The comparison of the two in terms of geometrical configu­
feasibility of the design. Although this is a relatively uncertain method ration and potential advantages of each is presented here. Fig. 12 shows
compared to traditional pure thermochemical hydrogen production, the the MSR reaction systems corresponding to different concentrators.
potential benefits are objective. By exploring the concentrator optimi­ Compared with PTSR systems, cylindrical concentrators can have better
zation, reactor design, operating conditions optimization, dynamic/ resistance to dust, wind, and rain because all components are assembled
transient characteristics research and cost analysis, we can further in a single cylinder [60]. Secondly, the STR has the advantage of good
improve and optimize the use of solar energy for thermochemical cat­ structural stability and easy operation because it is mounted near the
alytic processes more efficiently. ground. However, the PTSR system can be retrofitted on current CSP
The first problem to be solved is the influence of irradiation intensity projects, reducing some cost advantages when scaled up. Considering
on the stability of hydrogen production system. For concentrating solar the current cost of industrial blue H2 production ($1.2 to $2.8/kg) and
power system (CSP), the surface heat flux (profiled as surface temper­ the increase in the carbon tax, it is expected to accelerate the substitu­
ature) of the receiver generates obvious temperature gradient and tion of green H2 for other energy types. Green H2 production via STC
thermal stress with the change of irradiation [57,58]. The receiver reactions still has excellent potential for development until the cost is
working under non-uniform irradiation for a long time will produce comparable to or lower than that of blue one.
fatigue damage, structural deformation and even expiration [59]. The
measures to solve this problem are as described in the introduction. In 4. Conclusions
this paper, the STR surface temperature gradient is homogenized by
optimizing the structure of the concentrator and adjusting the relative The novel concentrated solar thermal collector system for MSR was
position. However, the structure of STR material and the temperature presented numerically to produce clean hydrogen. The photo-thermal-
characteristics under more irradiation conditions need further study, fluid-chemical conversion and coupling processes are considered to
especially the experimental verification. analyze the multi-physical field characteristics of this 3-dimensional
For the in-depth analysis of the cost, the ideal way is to compare it MSR-STR system and to optimize its methanol conversion and
with the cost of the PTSR system that has been implemented to confirm hydrogen production performance. Based on the time varying STR
the implementable economy of the system. Some results have been re­ concentration and heat collection efficiency, system geometry, STR inlet
ported for current solar-driven thermochemical hydrogen synthesis material ratio and temperature, reaction kinetics and Cu/ZnO/Al2O3
systems. Unfortunately, a cost analysis of PTSR systems has not been based catalyst properties are considered and discussed in detail. Based

Fig. 12. Comparison of novel MSR-STR system with PTSR system.

11
N. Zhao et al. Renewable Energy 222 (2024) 119952

on the results presented in this study, the following conclusions can be with high-proportion renewable energy, Energy (2023) 263, https://doi.org/
10.1016/j.energy.2022.125779.
drawn.
[5] S. Tang, J. Sun, H. Hong, Q. Liu, Solar fuel from photo-thermal catalytic reactions
with spectrum-selectivity: a review, Front. Energy 11 (2017) 437–451, https://doi.
(1) The uniformity of the energy flux density on the STR surface can org/10.1007/s11708-017-0509-z.
be optimized by adjusting the imaging distance of the STR, the [6] J.N. Hong, C.Y. Xu, B.W. Deng, Y. Gao, X. Zhu, X.H. Zhang, et al., Photothermal
chemistry based on solar energy: from synergistic effects to practical applications,
opening distance of the CPC, and the center distance between the Adv. Sci. 9 (2022) 26, https://doi.org/10.1002/advs.202103926.
STR and the CPC reflector. This adjustment is necessary consid­ [7] X.P. Tao, Y. Zhao, S.Y. Wang, C. Li, R.G. Li, Recent advances and perspectives for
ering the hydrogen yield and the safe operating range of the Cu/ solar-driven water splitting using particulate photocatalysts, Chem. Soc. Rev.
(2022) 48, https://doi.org/10.1039/d1cs01182k.
ZnO/Al2O3 catalyst. The adjusted MSR-STR system achieves a [8] N. Abas, E. Kalair, A. Kalair, Q. ul Hasan, N. Khan, Nature inspired artificial
maximum optical efficiency of 91 % and the flux density reaches photosynthesis technologies for hydrogen production: barriers and challenges, Int.
a preset value within 180◦ that can drive the MSR reaction. J. Hydrogen Energy 45 (2020) 20787–20799, https://doi.org/10.1016/j.
ijhydene.2019.12.010.
(2) The concentrating performance of the cylindrical Fresnel [9] S. Yoshino, A. Iwase, Y. Yamaguchi, T.M. Suzuki, T. Morikawa, A. Kudo,
concentrator is affected by the longitudinal incidence angle β and Photocatalytic CO2 reduction using water as an electron donor under visible light
tracking error angle γ. A single-axis tracker is sufficient to meet irradiation by Z-scheme and photoelectrochemical systems over (CuGa)0.5ZnS2 in
the presence of basic additives, J. Am. Chem. Soc. (2022), https://doi.org/
the required heat demand of the reaction for the MSR-STR sys­ 10.1021/jacs.1c12636.
tem. As the β increases, this MSR-STR exhibits a lower concen­ [10] Y. Zhang, G. Brooks, A. Rhamdhani, C. Guo, Review on solar thermochemical
trating efficiency. In addition, when the γ is larger than 20◦ , the processing for lunar applications and their heat transfer modeling methods, J. Heat
Tran. (2021) 143, https://doi.org/10.1115/1.4052080.
light divergence is severe, and the optical efficiency of the system
[11] W. Du, M.-T. Lee, Y. Wang, C. Zhao, G. Li, M. Li, Design of a solar-driven methanol
decreases significantly. steam reforming receiver/reactor with a thermal storage medium and its
(3) Higher reaction inlet temperature (513 K) and larger WFR (110 performance analysis, Int. J. Hydrogen Energy 45 (2020) 33076–33087, https://
kg s/mol) correspond to higher hydrogen yields (0.77 mol/s) for doi.org/10.1016/j.ijhydene.2020.09.045.
[12] R. Chein, Y.C. Chen, J.N. Chung, Axial heat conduction and heat supply effects on
the MSR-STR system under the control strategy of the Tf, max. In methanol-steam reforming performance in micro-scale reformers, Int. J. Heat Mass
addition, more steam concentrate will increase with increasing Tran. 55 (2012) 3029–3042, https://doi.org/10.1016/j.
SMR value, which will reverse the dilution of hydrogen concen­ ijheatmasstransfer.2012.02.022.
[13] R. Chein, Y.C. Chen, J.N. Chung, Numerical study of methanol-steam reforming
tration within the STR. and methanol-air catalytic combustion in annulus reactors for hydrogen
(4) The LCOH analysis show that the solar-driven MSR-STR system production, Appl. Energy 102 (2013) 1022–1034, https://doi.org/10.1016/j.
has an LCOH of $1.77/kg over a 10-year payback period, and this apenergy.2012.06.010.
[14] X. Pan, J. Qin, S. Zhang, W. Bao, Numerical analysis of heat transfer and flow
type of system shows potential for scale-up compared to the in­ characteristics of a precooler with a methanol cracking reaction, Case Stud. Therm.
dustrial cost of blue H2 production. In addition, due to the smaller Eng. 34 (2022) 102065, https://doi.org/10.1016/j.csite.2022.102065.
footprint, the system can be flexibly installed in urban areas, [15] C.F. Shih, T. Zhang, J.H. Li, C.L. Bai, Powering the future with liquid sunshine,
Joule 2 (2018) 1925–1949, https://doi.org/10.1016/j.joule.2018.08.016.
reducing the cost of feedstock and product transport on the one [16] Q. Liu, Y. Wang, J. Lei, H. Jin, Numerical investigation of the thermophysical
hand, and enhancing the portability of H2 production on the characteristics of the mid-and-low temperature solar receiver/reactor for hydrogen
other. production, Int. J. Heat Mass Tran. 97 (2016) 379–390, https://doi.org/10.1016/j.
ijheatmasstransfer.2016.02.012.
[17] D. Laudenschleger, H. Ruland, M. Muhler, Identifying the nature of the active sites
CRediT authorship contribution statement in methanol synthesis over Cu/ZnO/Al2O3 catalysts, Nat. Commun. 11 (2020)
3898, https://doi.org/10.1038/s41467-020-17631-5.
[18] D. Li, F. Xu, X. Tang, S. Dai, T. Pu, X. Liu, et al., Induced activation of the
Ning Zhao: Conceptualization, Data curation, Methodology, Writing commercial Cu/ZnO/Al2O3 catalyst for the steam reforming of methanol, Nat.
– original draft. Jiangjiang Wang: Methodology, Supervision, Writing – Catal. 5 (2022) 99–108, https://doi.org/10.1038/s41929-021-00729-4.
review & editing, Funding acquisition. Yuyang Tian: Investigation, [19] J. Kang, Y. Song, T. Kim, S. Kim, Recent trends in the development of reactor
systems for hydrogen production via methanol steam reforming, Int. J. Hydrogen
Software, Visualization. Zibo Yao: Visualization, Data curation. Suying
Energy 47 (2022) 3587–3610, https://doi.org/10.1016/j.ijhydene.2021.11.041.
Yan: Investigation, Data curation, Supervision. [20] F. Wang, L. Li, Y. Liu, Effects of flow and operation parameters on methanol steam
reforming in tube reactor heated by simulated waste heat, Int. J. Hydrogen Energy
42 (2017) 26270–26276, https://doi.org/10.1016/j.ijhydene.2017.09.002.
Declaration of competing interest [21] A.M. Ranjekar, G.D. Yadav, Steam reforming of methanol for hydrogen production:
a critical analysis of catalysis, processes, and scope, Ind. Eng. Chem. Res. 60 (2021)
89–113, https://doi.org/10.1021/acs.iecr.0c05041.
The authors declare that they have no known competing financial [22] Z.-D. Cheng, J.-J. Men, Y.-L. He, Y.B. Tao, Z. Ma, Comprehensive study on novel
interests or personal relationships that could have appeared to influence parabolic trough solar receiver-reactors of gradually-varied porosity catalyst beds
the work reported in this paper. for hydrogen production, Renew. Energy 143 (2019) 1766–1781, https://doi.org/
10.1016/j.renene.2019.05.137.
[23] M. Liu, Y. Shi, N. Cai, Modeling of packed bed methanol steam reformer integrated
Acknowledgements with tubular high temperature proton exchange membrane fuel cell, J. Therm. Sci.
32 (2023) 81–92, https://doi.org/10.1007/s11630-022-1764-9.
[24] X.-Y. Tang, W.-W. Yang, X. Ma, X.E. Cao, An integrated modeling method for
This research has been supported by the National Natural Science membrane reactors and optimization study of operating conditions, Energy 268
Foundation of China (Grant No. 52276007) and the Fundamental (2023) 126730, https://doi.org/10.1016/j.energy.2023.126730.
[25] T. Zhang, X.-Y. Tang, W.-W. Yang, X. Ma, Comprehensive performance study on
Research Funds for the Central Universities (Grant No.2023JC010).
reflux solar methanol steam reforming reactor for hydrogen production, Int. J.
Hydrogen Energy 48 (2023) 879–893, https://doi.org/10.1016/j.
References ijhydene.2022.10.002.
[26] Y.Y. Ling, W.J. Li, J. Jin, Y.H. Yu, Y. Hao, H.G. Jin, A spectral-splitting
photovoltaic-thermochemical system for energy storage and solar power
[1] H. Song, S. Luo, H. Huang, B. Deng, J. Ye, Solar-driven hydrogen production:
generation, Appl. Energy 260 (2020), https://doi.org/10.1016/j.
recent advances, challenges, and future perspectives, ACS Energy Lett. 7 (2022)
apenergy.2019.113631.
1043–1065, https://doi.org/10.1021/acsenergylett.1c02591.
[27] T. Zhu, Q. Li, A. Yu, Analysis of the solar spectrum allocation in a spectral-splitting
[2] X. Shi, X. Zhao, F. Wang, Z. Cheng, Y. Dong, J. Xu, Improving overall heat transfer
photovoltaic-thermochemical hybrid system, Sol. Energy 232 (2022) 63–72,
performance of parabolic trough solar receiver by helically convex absorber tube,
https://doi.org/10.1016/j.solener.2021.12.022.
Appl. Therm. Eng. 213 (2022), https://doi.org/10.1016/j.
[28] Y.-L. He, K. Wang, Y. Qiu, B.-C. Du, Q. Liang, S. Du, Review of the solar flux
applthermaleng.2022.118690.
distribution in concentrated solar power: non-uniform features, challenges, and
[3] M. Martino, C. Ruocco, E. Meloni, P. Pullumbi, V. Palma, Main hydrogen
solutions, Appl. Therm. Eng. 149 (2019) 448–474, https://doi.org/10.1016/j.
production processes: an overview, Catalysts 11 (2021), https://doi.org/10.3390/
applthermaleng.2018.12.006.
catal11050547.
[29] Z.D. Cheng, Y.K. Leng, J.J. Men, Y.L. He, Numerical study on a novel parabolic
[4] R. Yan, J. Wang, S. Huo, Y. Qin, J. Zhang, S. Tang, et al., Flexibility improvement
trough solar receiver-reactor and a new control strategy for continuous and
and stochastic multi-scenario hybrid optimization for an integrated energy system

12
N. Zhao et al. Renewable Energy 222 (2024) 119952

efficient hydrogen production, Appl. Energy 261 (2020) 13, https://doi.org/ [46] B.A. Peppley, J.C. Amphlett, L.M. Kearns, R.F. Mann, Methanol–steam reforming
10.1016/j.apenergy.2019.114444. on Cu/ZnO/Al2O3 catalysts. Part 2. A comprehensive kinetic model, Appl. Catal.
[30] X.X. Yu, J. Zeng, Y.M. Xuan, A direct solar photo-thermal conversion of methanol Gen. 179 (1999) 31–49, https://doi.org/10.1016/S0926-860X(98)00299-3.
into hydrogen, Energy Technol. 7 (2019) 9, https://doi.org/10.1002/ [47] F. Wang, J. Tan, L. Ma, Y. Leng, Effects of key factors on solar aided methane steam
ente.201900299. reforming in porous medium thermochemical reactor, Energy Convers. Manag. 103
[31] T. Zhu, Q. Li, Y. Xuan, D. Liu, H. Hong, Performance investigation of a hybrid (2015) 419–430, https://doi.org/10.1016/j.enconman.2015.06.049.
photovoltaics and mid-temperature methanol thermochemistry system, Appl. [48] P. Harold M, B. Nair, G. Kolios, Hydrogen generation in a Pd membrane fuel
Energy 256 (2019), https://doi.org/10.1016/j.apenergy.2019.113908. processor: assessment of methanol-based reaction systems, Chem. Eng. Sci. 58
[32] N. Bushra, T. Hartmann, A review of state-of-the-art reflective two-stage solar (2003) 2551–2571, https://doi.org/10.1016/S0009-2509(03)00105-2.
concentrators: technology categorization and research trends, Renew. Sustain. [49] Z.D. Cheng, Y.L. He, F.Q. Cui, R.J. Xu, Y.B. Tao, Numerical simulation of a
Energy Rev. 114 (2019) 109307, https://doi.org/10.1016/j.rser.2019.109307. parabolic trough solar collector with nonuniform solar flux conditions by coupling
[33] R. Ma, J. Sun, D.H. Li, J.J. Wei, Review of synergistic photo-thermo-catalysis: FVM and MCRT method, Sol. Energy 86 (2012) 1770–1784, https://doi.org/
mechanisms, materials and applications, Int. J. Hydrogen Energy 45 (2020) 10.1016/j.solener.2012.02.039.
30288–30324, https://doi.org/10.1016/j.ijhydene.2020.08.127. [50] Z.D. Cheng, J.J. Men, X.R. Zhao, Y.L. He, Y.B. Tao, A comprehensive study on
[34] N. Zhao, S. Yan, N. Zhang, X. Zhao, Impacts of seasonal dust accumulation on a parabolic trough solar receiver-reactors of methanol-steam reforming reaction for
point-focused Fresnel high-concentration photovoltaic/thermal system, Renew. hydrogen production, Energy Convers. Manag. 186 (2019) 278–292, https://doi.
Energy 191 (2022) 732–746, https://doi.org/10.1016/j.renene.2022.04.039. org/10.1016/j.enconman.2019.02.068.
[35] M.K. Sharma, J. Bhattacharya, Deciding between concentrated and non- [51] Z.-D. Cheng, J.-J. Men, X.-R. Zhao, Y.-L. He, Y.-B. Tao, A comprehensive study on
concentrated photovoltaic systems via direct comparison of experiment with opto- parabolic trough solar receiver-reactors of methanol-steam reforming reaction for
thermal computation, Renew. Energy 178 (2021) 1084–1096, https://doi.org/ hydrogen production, Energy Convers. Manag. 186 (2019) 278–292, https://doi.
10.1016/j.renene.2021.06.128. org/10.1016/j.enconman.2019.02.068.
[36] Z.W. Zhang, J. Sun, D.H. Li, R. Ma, J.J. Wei, Proposal of a novel modular photo- [52] B.A. Peppley, A Comprehensive Kinetic Model of Methanol-Steam Reforming on
thermo-reactor system for cascaded hydrogen production from methanol steam Cu/ZnO/Al (2) O (3) Catalyst, 1998.
reforming, Energy Convers. Manag. 256 (2022), https://doi.org/10.1016/j. [53] F.J. Keil, Complexities in modeling of heterogeneous catalytic reactions, Comput.
enconman.2022.115390. Math. Appl. 65 (2013) 1674–1697, https://doi.org/10.1016/j.
[37] R. Leutz, A. Suzuki, A. Akisawa, T. Kashiwagi, Design of a nonimaging Fresnel lens camwa.2012.11.023.
for solar concentrators, Sol. Energy 65 (1999) 379–387, https://doi.org/10.1016/ [54] J.R. Hendry, J.G.M. Lee, M.J. Battrum, CFD model of fluid flow and particle
S0038-092X(98)00143-1. deposition during cryogenic condensation, Chem. Eng. Res. Des. 143 (2019)
[38] H. Zheng, C. Feng, Y. Su, J. Dai, X. Ma, Design and experimental analysis of a 201–214, https://doi.org/10.1016/j.cherd.2019.01.016.
cylindrical compound Fresnel solar concentrator, Sol. Energy 107 (2014) 26–37, [55] R.-Y. Chein, Y.-C. Chen, H.-J. Zhu, J.N. Chung, Numerical simulation of flow
https://doi.org/10.1016/j.solener.2014.05.010. disturbance and heat transfer effects on the methanol-steam reforming in miniature
[39] X. Ma, H. Zheng, S. Liu, Optimization on a cylindrical Fresnel lens and its annulus type reformers, Energy Fuels 26 (2012) 1202–1213, https://doi.org/
validation in a medium-temperature solar steam generation system, Renew. Energy 10.1021/ef201498t.
134 (2019) 1332–1343, https://doi.org/10.1016/j.renene.2018.08.075. [56] P. Kurup, C.S. Turchi, Parabolic Trough Collector Cost Update for the System
[40] X. Ma, R. Jin, S. Liang, S. Liu, H. Zheng, Analysis on an optimal transmittance of Advisor Model (SAM), National Renewable Energy Laboratory (NREL), 2015.
Fresnel lens as solar concentrator, Sol. Energy 207 (2020) 22–31, https://doi.org/ [57] M. Perrero, D. Papurello, Solar disc concentrator: material selection for the
10.1016/j.solener.2020.06.071. receiver, Energies 16 (2023) 6870.
[41] TracePro User Manual. Lambda Research Corporation. [58] H. Shatnawi, C.W. Lim, F.B. Ismail, A. Aldossary, An optimisation study of a solar
[42] ANSYS Fluent Users Guide 2020R2. ANSYS Inc.. tower receiver: the influence of geometry and material, heat flux, and heat transfer
[43] R. Jeschar, Druckverlust in Mehrkornschüttungen aus Kugeln, Arch. für das fluid on thermal and mechanical performance, Heliyon 7 (2021) e07489, https://
Eisenhuttenwes. 35 (1964) 91–108, https://doi.org/10.1002/srin.196402300. doi.org/10.1016/j.heliyon.2021.e07489.
[44] B.E. Poling, J.M. Prausnitz, J. Paul, C. O, The Properties of Gases and Liquids, fifth [59] Y. Luo, G. Li, Z. Wang, T. Lu, Analysis of temperature and thermal stress for a solar
ed., McGraw-Hill, New York, 2001. power tower molten salt receiver under multi-source uncertainties, Appl. Sci. 12
[45] L.-T. Zhu, W.-Y. Ma, Z.-H. Luo, Influence of distributed pore size and porosity on (2022) 10740.
MTO catalyst particle performance: modeling and simulation, Chem. Eng. Res. Des. [60] N. Zhao, Z. Wu, H. Gao, S. Yan, X. Zhao, N. Zhang, et al., Experimental analysis of
137 (2018) 141–153, https://doi.org/10.1016/j.cherd.2018.07.005. the optical loss of a dusty Fresnel lens with a novel solar flux test system, Sustain.
Energy Technol. Assessments 48 (2021), https://doi.org/10.1016/j.
seta.2021.101656.

13