reactor for hydrogen production

Solar Energy 134 (2016) 273–283

Contents lists available at ScienceDirect

Solar Energy journal homepage: www.elsevier.com/locate/solener

A three-dimensional simulation of a mid-and-low temperature solar receiver/reactor for hydrogen production Yanjuan Wang a,b,c, Qibin Liu b,c,⇑, Jing Lei a, Hongguang Jin b,c a

School of Energy, Power and Mechanical Engineering, North China Electric Power University, Changping District, Beijing 102206, China Institute of Engineering Thermophysics, Chinese Academy of Sciences, Beijing 100190, China c University of Chinese Academy of Sciences, Beijing 100049, China b

a r t i c l e

i n f o

Article history: Received 30 October 2015 Received in revised form 3 May 2016 Accepted 5 May 2016

Keywords: Solar hydrogen production Solar receiver/reactor Methanol steam reforming Solar fuels

a b s t r a c t The solar receiver/reactor is a key component that influences the conversion efficiency in the solar thermochemical process. A thermochemical solar reactor/receiver consisting of a porous Cu/ZnO/Al2O3 catalyst bed is studied in this paper. A three-dimensional thermochemical coupling model that incorporates the fluid flowing through the porous catalyst bed and energy conservation equations coupling the radiation/convection/conduction heat transfer with the reaction kinetics is proposed to investigate the performances of the receiver/reactor. The factors of influencing the hydrogen production and the temperature distribution, including the mole ratio of water/methanol, the solar radiation and the inlet temperature, are numerically investigated. Numerical simulation results indicate that the deactivation of the catalyst may appear near the receiver/reactor tube wall. The methanol conversion decreases with the increase of the methanol feeding rate, and the low inlet methanol feeding rate should be avoided for the protection of the catalyst bed. A new solar receiver/reactor is proposed by changing the aperture width along the flow direction to make the concentrated solar energy level match the chemical reaction. Compared with traditional solar receivers/reactors, the thermochemical efficiency can be increased by 3% points. The research findings will pave the way for the future development of the mid-and-low temperature solar receiver/reactor. Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction Solar energy is considered to be a promising energy in the 21th century due to its infinite reserves and cleanness nature. It may be one of most effective routes to solve the energy problems caused by the exploitation and utilization of the fossil fuels. Among the three main solar thermal technologies, i.e., the parabolic trough, the central receiver and the parabolic dish, the parabolic trough solar technology is the most proven and cost-effective, largescale solar power technology available today (Price et al., 2002; Xu et al., 2015). Hydrogen, as a fuel, likely becomes one of most promising energy carriers in the near future for environmentally benign and sustainable development. The major benefit of using hydrogen as an energy carrier is that it only produces water and liberates large amounts of energy per unit weight as be combusted. With the increasing demand of hydrogen, it is anticipated that the role of ⇑ Corresponding author at: Institute of Engineering Thermophysics, Chinese Academy of Sciences, Beijing 100190, China. E-mail address: [email protected] (Q. Liu). http://dx.doi.org/10.1016/j.solener.2016.05.003 0038-092X/Ó 2016 Elsevier Ltd. All rights reserved.

hydrogen will become more significant (Abuiadala and Dincer, 2012; Gadalla et al., 2010; Muradov and Veziroglu, 2008). Recently, hydrogen production by solar energy has received increasing attentions. Many potential methods have been exploited. Chueh et al. proposed a simple and scalable reactor using the porous ceria directly exposed to the concentrated solar radiation, and thus the high-temperature heat can be transformed to the reaction sites. They studied the feasibility of a solar-driven thermochemical cycle for the dissociation of H2O and CO2 using the nonstoichiometric ceria (Chueh et al., 2010). The thermochemical hydrogen production from a two-step solar-driven watersplitting cycle based on cerium oxides and solar hydrogen production from the thermal splitting of methane in a high temperature solar chemical reactor were studied by Stephane Abanades (Abanades and Flamant, 2006a,b). Zamfirescu and Dincer (2014) developed a novel integrated system that converts solar radiation into hydrogen by combining photocatalytic reactor, photovoltaics, thermal engine and chemical energy storage for solar energy harvesting. It was shown that the annual average factor of the light absorption of the novel system is increased by 9% points. Furthermore, the overall exergy efficiency of the hydrogen production is

274

Y. Wang et al. / Solar Energy 134 (2016) 273–283

Nomenclature dp DTi Dij M nCH3 OH nun;CH3 OH p Q Q so
ch r R T u xCH3 OH Dr H

equivalent diameter (m) the generalized thermal diffusion coefficient (kg/(m s)) the ij component of the multi-component diffusivity coefficient (m2/s) molar mass of species mole feeding rate of methanol non-reacted methanol pressure (Pa) heat source (W/m3) he solar thermal energy converted into the chemical energy reaction rate (mol/(m3 s)) universal gas constant temperature (K) velocity (m/s) methanol conversion enthalpy of reaction (J/mol)

increased by 40%, as compared with a conventional tower system that runs an electrolyzer. The solar hydrogen production by the thermal decomposition of natural gas using a vortex-flow reactor was proposed by Hirsch (Hirsch and Steinfeld, 2004). Thermal characterizations of a cavity receiver for the hydrogen production by the thermochemical cycles operating at a moderate temperature had been investigated by Lanchi et al. (2013), whereas other successful examples can be found elsewhere for other solar hydrogen production techniques (Steinfeld and Meier, 2004; Steinfeld, 2005; Segal and Epstein, 2003; Dufour et al., 2009). The above-mentioned hydrogen production methods often need to concentrate the solar thermal energy above 800 K to provide the necessary reaction heat, which leads to technological difficulties on using the solar radiation as a driving energy for reactions. More recently, based on the parabolic trough technology, an effective hydrogen production approach via the integrating utilization of the middle temperature solar thermal energy and methanol was proposed by Jin and his group (Jin et al., 2007; Hong et al., 2005; Liu et al., 2009, 2010). Compared with other hydrocarbons, methanol has certain advantages, such as the relatively low reforming temperature (423–573 K) and easy implementation. Many benefits can be achieved in the process, e.g., ease implementation, accurate tracking of solar concentrators, low-cost investments, etc. Jin et al. (2007) developed an original mid-and-low temperature solar receiver/reactor prototype. The mid-and-low temperature solar receiver/reactor is a key component in the mid-and-low temperature solar thermochemical process, and directly influences the conversion efficiency of solar thermal energy into chemical energy, and thus the investigations of the characteristics of the receiver/reactor are crucial for practical applications. Liu et al. (2010) investigated a novel solar thermochemical receiver/reactor, and developed a non-isothermal model to analyze the performances of the mid-and-low temperature solar receiver/reactor.Hou et al. (2007) developed a procedure to analyze the performances of the non-isothermal solar reactors for the methanol decomposition. However, the considerations of the influences of the non-uniform three-dimensional distribution of the solar flux on the solar hydrogen production are absent. The understanding of the complicated physical and chemical mechanism of the hydrogen production by the steam methanol reforming method is beneficial for the efficient utilization of mid-and-low temperature solar thermal energy. In this paper, naturally, we employ the numerical simulation approach to investigate the performances of the receiver/reactor under non-uniform

Subscript f gaseous phases g gas mixture s catalyst bed Greek symbols dynamic viscosity (kg/(m s)) hydraulic permeability (m2) porosity ksr effective thermal conductivity of the catalyst bed (W/(m K)) ks thermal conductivities of catalyst particle (W/(m K)) kg thermal conductivities of gas mixture (W/(m K)) xi mass fraction of the i-th gas mole fraction of the j-th gas fj gso
ch thermochemical efficiency

g j e

distribution of the solar flux, and the influences of the key operating parameters on the thermal, fluid and chemical characteristics of the solar receiver/reactor are revealed. The main contributions are summarized as: (1) The solar receiver/reactor consists of a porous catalyst bed made of Cu/ZnO/Al2O3 (see Fig. 1 for more details) where the endothermic methanol steam reforming reaction occurs. A multiphysics coupling model of the receiver/reactor is proposed to model the steam methanol reforming process, which incorporates mass, momentum and energy conservation equations as well as the methanol steam reforming reaction. The solar flux profile is calculated by the solar ray-tracing method. The thermal and chemical characteristics of the mid-and-low temperature solar receiver/reactor are revealed. (2) The influences of the key operating parameters on the performances of the mid-and-low temperature solar receiver/ reactor are investigated, including solar radiation, inlet temperature of reactants, mole flow rate of reactants and mole ratio of water/methanol on the deactivation of the catalyst and hydrogen production. (3) Due to the characteristics of the chemical reaction of the methanol steam reforming, we find that the required reaction heat is large at the beginning part of the chemical reaction, and becomes very small when the reaction is almost reacted. Thus, a new solar receiver/reactor is proposed by changing the aperture width along the flow direction to well match the concentrated solar energy level with the chemical reaction. The rest of the paper is structured as follows. The mid-and-low temperature solar receiver/reactor model and the performance simulations of the mid-and-low temperature solar receiver/reactor are implemented in Section 2. The numerical results and detailed discussion on the results are presented in Section 3. Finally, Section 4 summarizes the main conclusions. 2. Mid-and-low temperature solar receiver/reactor model 2.1. Configuration of the solar receiver/reactor The methanol steam reforming requires energy input with a range of about 473–573 K, which can be supplied by the solar

275

Y. Wang et al. / Solar Energy 134 (2016) 273–283

Fig. 1. Cu/ZnO/Al2O3 catalyst.

Glass tube

Annulus space

Catalyst bed

Receiver/reactor tube Fig. 3. Configuration of the mid-and-low-temperature receiver/reactor.

Table 1 Geometric structure parameters of the solar receiver/reactor.

Fig. 2. Photograph of the solar receiver/reactor prototype.

energy collected by parabolic trough collectors. Water and methanol are mixed in stoichiometric amounts and enter through the inlet of the solar receiver/reactor. In the solar receiver/reactor, the endothermic reaction is driven by the concentrated solar energy. The reaction of water and methanol mainly produces hydrogen and carbon dioxide:

CH3 OH þ H2 O ! CO2 þ 3H2 ;

Dr H298K ¼ 50:7 kJ=mol

The experiments on hydrogen production using mid-and-low temperature solar energy as the driving source with methanol steam reforming were carried out in a modified solar receiver/reactors shown in Fig. 2 (Hong et al., 2009), the solar receiver/reactor consisted of the one-tracking parabolic trough concentrators and a fixed-bed receiver/reactor, which was positioned along the focal line of the parabolic trough concentrator. The parabolic trough solar collector was used to concentrate the solar energy for the methanol steam reforming. The solar receiver featured a tubular reactor, 31 mm in diameter and 4 m in length, laden with catalysts of Cu/ZnO/Al2O3 (see Fig. 1 for more details). A commercial Cu/ZnO/Al2O3 catalyst was used in this paper. We listed the catalyst properties provided by the manufacturer in Table 2. The schematic configuration of the mid-and-low temperature solar receiver/reactor was shown in Fig. 3, and the geometric structure parameters of the receiver/reactor were listed in Table 1.

Items

Parameter

Aperture width Aperture length Focal length Inner diameter of reactor Outer diameter of reactor Inner diameter of glass tube Outer diameter of glass tube Transmittance of glass tube Coating absorbance

2.5 m 4m 2.25 m 31 mm 35 mm 51 mm 54 mm 0.95 0.95

Table 2 Typical physical properties of the catalysts particles and catalysts bed. Items

Parameter

Catalysts particles CuO ZnO Al2O3 Density

40 wt% 40 wt% 20 wt% 5767 kg/m3

Catalyst bed Pellet size Porosity Equivalent diameter Specific surface area

U5  2.5 mm 0.49 2.44 mm 102 m2/g

2.2. Simulations of the solar receiver/reactor A three-dimensional numerical simulation is implemented to analyze the process of the solar radiation converted to the chemical energy in a mid-and-low temperature receiver/reactor, including the solar radiation absorption in the receiver/reactor tube, the heat transfer between the catalyst bed and the receiver/reactor

276

Y. Wang et al. / Solar Energy 134 (2016) 273–283

70

Local concentration ratio

60

Fig. 4. Coupling model of the mid-and-low temperature solar receiver/reactor.

tube, the conduction and convection heat transfer and chemical reaction in the catalyst bed, the conduction heat transfer through the receiver/reactor tube and the glass tube, the heat transfer from the receiver/reactor tube to the glass tube and the heat transfer from the glass tube to the atmosphere, as shown in Fig. 4.

50 40 30 20 10 0 0

45

90

135

180

225

270

315

360

Circumferential angle/degree Fig. 5. Local concentration ratio distribution on the outer surface of the receiver/ reactor.

2.3. Heat transfer of the receiver/reactor tube and glass tube In real-world applications, it is crucial to acquire the accurate solar flux distributions in a PTC system for the three-dimensional numerical model. In this paper the ray-tracing method is employed to simulate the focus process of the sun rays due to its easy numerical implementation and high computational efficiency. The SolTrace software is employed to compute the non-uniform solar energy flux distribution on the outer surface of the receiver/reactor tube (Wendelin, 2003; Wang et al., 2015). Fig. 5 shows the local concentration ratio distribution on the outer surface of the receiver/reactor tube generated by SolTrace. The solar energy flux distribution is non-uniform in the circumferential direction and it is symmetrical approximately. The non-uniform solar energy flux distribution brings circumferential mechanical stress in the receiver/reactor tube leading to the deformation of the receiver/reactor tube. Then, the result calculated by the ray-tracing method is used as the boundary condition of the receiver/reactor tube’s outer surface. The heat transfer in the receiver/reactor tube and the glass tube is the heat conduction. The boundary conditions can be summarized as follow: thermal wall function at the receiver/reactor tube’s inner surface; radiation and non-uniform solar flux distribution at the receiver/reactor tube’s outer surface; radiation at the glass tube’s inner surface; mixed boundary of radiation and convection at the glass tube’s outer surface; thermal insulation at the tube ends. It is noticeable that the radiation heat transfer occurs because of the temperature difference between the outer receiver/reactor tube surface and the inner glass tube wall. The glass is opaque to infrared radiation (Forristall, 2003), thus, in this paper the radiation heat transfer calculation is simplified according to the assumption that the glass tube is opaque to infrared radiation and gray surfaces. The radiation is described by the surface-to-surface radiation. At the outer surfaces of the glass tube, radiation is described by the surface-to-ambient radiation, without the consideration of the reflected radiation from the surroundings. 2.4. Fluid flow, heat and mass transfer of the porous catalyst bed The flow of the gaseous species through the porous catalyst bed is described by the Darcy’s law, the flow variables and fluid properties are averaged over a control volume surrounding a point. According to the Darcy’s law, the velocity distributions are determined by the pressure gradient, the fluid viscosity and the structure of the porous medium, which can be formulated as:




u¼




j @p @p @p ; þ þ g @x @y @z

ð1Þ

where g stands for the dynamic viscosity of the fluid (kg/(m s)); p represents the pressure (Pa); u stands for the velocity (m/s) and the hydraulic permeability j (m2) is described by Fogler (1986):

dp e3 2

j¼

150ð1
eÞ2

;

ð2Þ

where e represents the porosity of the catalyst bed and dp is the equivalent diameter of the catalyst particles. In the porous catalyst bed, the heat transfer (including conduction and convection) and the chemical reaction occur simultaneously. Assuming that the porous catalyst bed is homogeneous and isotropic (Kanouff et al., 2013), and thus the energy transport equation can be described as


ksr


 @ @T s @T s þ ðqC p Þf ui ¼ Q; @xi @xi @xi

ð3Þ

Q ¼ Dr H  r;

ð4Þ

ksr ¼ ekg þ ð1
eÞks ;

ð5Þ

where subscript f denotes gaseous phases; T s represents the temperature (K) of the catalyst bed; ksr stands for the effective thermal conductivity (W/(m K)) of the catalyst bed; ks and kg are the thermal conductivities of catalyst particle and gas mixture, respectively; Q represents a heat source (W/m3) due to the chemical reaction; u is the fluid velocity (m/s); Dr H defines the enthalpy of reaction (J/mol) and r represents the reaction rate (mol/(m3 s)). The kinetics of the methanol steam reforming process using Cu/ ZnO/Al2O3 has been investigated by many researchers. According to the experiments (Liu et al., 2009), there is a little amount of CO in the gas products, and the composition of the gas products indicate that H2 and CO2 are produced at an approximate ratio of 3:1, and thus an overall kinetic model of the methanol steam reforming is used (Jiang et al., 1993), which can be formulated as: rR ¼

T T 3 kR K CH3 Oð1Þ ðpCH3 OH =p1=2 H2 Þð1
pH2 pCO2 =kR pCH3 OH pH2 O ÞC S1 C S1a

ð1 þ K CH3 Oð1Þ ðpCH3 OH =pH2 Þ þ K HCOOð1Þ pCO2 pH2 þ K OHð1Þ ðpH2 O =pH2 ÞÞð1 þ K H1a pH2 Þ 1=2

1=2

1=2

1=2

1=2

;

ð6Þ

The parameters in Eq. (6) can be calculated by Jiang et al. (1993):

Y. Wang et al. / Solar Energy 134 (2016) 273–283



ER 1 kR ¼ kR exp RT K CH3 Oð1Þ

DSCH3 Oð1Þ

¼ exp

DHCH3 Oð1Þ

ð7Þ

the hydrogen production, and thus these factors are numerically investigated in this section.

ð8Þ

3.1. Temperature distributions of the solar receiver/reactor

ð9Þ

We simulated a typical working condition to investigate the performances of the solar thermochemical receiver/reactor, and it can be outlined as follows: the direct normal irradiation (DNI) is 800 W/m2, the mole flow rate of methanol is 0.06 mol/s, and the inlet temperature of reactants are 473.2 K and the mole ratio of water/methanol is fixed at 1:1. The temperature distributions of the solar receiver/reactor play an important role in the hydrogen production, and will directly affect the collector efficiency and the chemical reaction rate. Fig. 6 shows the temperature distributions of the catalyst bed. We can find from Fig. 6(a) that the temperature distribution of the catalyst bed is non-uniform in both the main flow direction and the radial direction, and the average temperature of the catalyst bed is about 502.5 K. To investigate the temperature distribution thoroughly, two cross sections x = 2 m and y = 0 m are chosen. As shown in Fig. 6(b) and (c), the temperature distributions in the radial directions are distinctly different, and the temperatures near the receiver/reactor tube wall are higher, especially the temperature at the bottom part of the receiver/reactor tube, the crosssectional temperature difference of the catalyst bed is 31.7 K, which is mainly derived from the non-uniform solar flux distribution that is higher at the bottom. The temperature of the catalyst bed increases from 462.1 K to 607.1 K along the main flow direction, owing to the fact that the working mediums are heated by the solar flux along the flow direction. According to (CN-31, 2012), the deactivation is prone to happen when the temperature of the catalyst bed is above 573.2 K which appears near the receiver/reactor tube wall in the latter part of the catalyst bed. In order to protect the catalyst bed, a high temperature of the catalyst bed should be avoided. Figs. 7 and 8 show the temperature distributions of the receiver/ reactor tube and glass tube, respectively. We can observe that the temperature become higher along the main flow direction. The average temperature of the receiver/reactor tube is about 530.4 K. The cross section temperature profiles of the receiver/ reactor tube are symmetrical approximately, like the local concentration ratio distribution (see Fig. 5), the maximum of the temperature difference of the receiver/reactor tube in the x = 2 m in the cross section is 69.1 K, which can bring circumferential mechanical stress in the receiver tubes, leading to the deformation of the receiver/reactor tube. The temperatures of the glass tube are in the range of 300.6–331.6 K, which are much lower than that of the receiver/reactor tube, leading to the reduction of the heat loss, especially the radiation heat loss. In this case, the convective heat loss is about 31.0 W/m, and the radiation heat loss is about 17.1 W/m. Fig. 9 shows the methanol conversion ratio along the solar receiver/reactor. In this paper, the methanol conversion ratio can be computed by:

!

þ R RT
  DSHCOOð1Þ DHHCOOð1Þ K HCOOð1Þ ¼ exp þ R RT
 DSH1a DHH1a þ K H1a ¼ exp R RT

ð10Þ

where r R is the reaction rate of the methanol; R represents the universal gas constant; T stands for the temperature of the solar receiver/reactor, pi defines the partial pressure of component i. The parameters of Eqs. (6)–(10) are presented in Table 3. At present, the Maxwell–Stefan equation has been widely used in various fields, such as heat and mass transfer process (Krishna and Wesselingh, 1997), mass transport with chemical reactions (Kuhn et al., 2009), mass and energy transport in a porous media (Krishna and van Baten, 2005), and the diffusion of charged particles (Silva and Lito, 2007). The theory model widely used is the generalized Maxwell–Stefan equation proposed by Amsterdam (Krishna, 1990). In this paper, the mass transport with the chemical reaction in the porous catalyst bed is described by the Maxwell–Stefan diffusion:

!
 rp T rT f ¼ Ri ; Dij rf j þ ðf j
xj Þ
Di r  qxi~ u
qxi p T j¼1 n X

ð11Þ where q is the density of the gas (kg/m3); xi represents the mass fraction of the i-th gas; f j stands for the mole fraction of the j-th gas; DTi denotes the generalized thermal diffusion coefficient kg/(m s); Ri is the reaction rate (kg/(m3 s)); Dij defines the ij component of the multi-component diffusivity coefficient (m2/s) which was studied by Fuller et al. Poling et al. (2001):

Dij ¼

0:00143T 1:75 !1=3 !1=3 32 ; X X 1=2 5 PMij 4 þ

ð12Þ

2

v




1 1 M ij ¼ 2 þ Mi Mj

i


1

v

j

ð13Þ

where Mi is the molar mass of species i; Mj represents the molar mass of species j, g mol
1; p stands for pressure, bar. 3. Results and discussion Since the solar radiation, the mole flow feeding rates, and the inlet temperature of the reactants directly influence the chemical reaction, the heat transfer and the fluid flow characteristics of the mid-and-low temperature solar thermochemical system for

xCH3 OH ¼

Table 3 Parameters of the kinetic model of methanol-steam reforming. Rate constant or equilibrium constant

DSi ðJ mol
1 K
1 Þ or 1

ki ðm2 s
1 mol kR ðm2 s
1 mol K CH

ð1Þ

3O

ðbar


0:5

K OHð1Þ ðbar


1


0:5

Þ

Þ


0:5 Þ K H1a ðbar
0:5  Þ K HCOOð1Þ ðbar

Þ


1

Þ

DHi or EðkJ mol

7.4E+14

102.8


41.8


20.0


44.6


20.0


100.8


50.0

197.2

100.0


1

277

Þ

nCH3 OH
nun;CH3 OH ; nCH3 OH

ð14Þ

where nCH3 OH is the mole feeding rate of methanol and nun;CH3 OH represents the non-reacted methanol. The average methanol conversion has the highest value on the fluid outlet surface, and the methanol conversion increases along the main flow direction due to the endothermic steam methanol reforming reaction. For the typical work condition, the DNI is 800 W/m2, the mole flow rate of methanol is 0.06 mol/s, and the inlet temperature of reactants is 473.2 K, the mole ratio of water/methanol is fixed at 1:1, and the methanol conversion at the reactor exit can reach 97.6%.

278

Y. Wang et al. / Solar Energy 134 (2016) 273–283

(b) x=2 m

(a) Temperature distributions

(c) y=0 m Fig. 6. Temperature distributions of the catalyst bed.

(a)

(b) x=2m

Fig. 7. Temperature distributions of the receiver/reactor tube.

Fig. 8. Temperature distributions of the glass tube.

Fig. 9. Methanol conversion in the catalyst bed.

Fig. 10(a) and (b) presents the CH3OH and H2O mole fraction distributions on the porous catalyst bed, respectively. The mole fractions of CH3OH and H2O have the highest value of 0.50 on the entrance surface, and the CH3OH mole fraction as well as the

H2O mole fraction decrease along the main flow direction due to the endothermic steam methanol reforming reaction, furthermore, the CH3OH and H2O mole fraction near the tube wall in the latter part of the solar receiver/reactor is almost reacted.

Y. Wang et al. / Solar Energy 134 (2016) 273–283

(a) CH3OH mole fraction distribution

279

(b) H2O mole fraction distribution

Fig. 10. Mole fraction distributions along the porous catalyst bed.

increase of the methanol conversion. With the increase of the methanol feeding rate, the residence time of the reactant gases becomes smaller, and the methanol conversion is therefore decreased. Fig. 12 presents the influence of methanol feeding rate and the inlet temperature of reactants on the outlet temperature of the solar receiver/reactor. As can be expected, the outlet temperature increases with the rising inlet temperature at a fixed methanol feeding rate. Under a same methanol feeding rate of 0.07 mol/s, the outlet temperature are 514.2 K, 523.1 K, 532.4 K and 549.8 K respectively. The outlet temperature decreases with the increase of methanol feeding rate at a fixed inlet temperature. More methanol feeding rate can absorb more solar energy, so the outlet temperature will decrease. When the reaction rate is lower than 0.07 mol/s, the outlet temperature becomes higher than 553.2 K and increases rapidly with the decrease of the methanol feeding rate. In order to protect the catalyst bed, the low feeding rate of reactants should be avoided. Fig. 11. Variations of methanol conversion with methanol feeding rate and inlet temperature.

3.3. Influence of solar radiation and mole ratio of water/methanol on hydrogen production 3.2. Role of the inlet temperature of reactants and mole flow rate of reactants in methanol conversion and the deactivation of the catalyst The reactants entered the solar receiver/reactor after preheated by the solar collectors or other kinds of energy. The inlet temperature of the reactants directly influences the initial reaction rate. Fig. 11 shows the correlations between the methanol conversion and the mole flow rate of methanol and the inlet temperature of reactants. The mole ratio of water/methanol is fixed at 1:1. Under the same solar radiation, i.e., 800 W m2, four different inlet temperatures, i.e., 373.2 K, 433.2 K, 473.2 K and 423.2 K, are selected for simulations. We can find that the methanol conversion increases with the augmentation of the inlet temperature for a fixed mole flow rate of methanol. For example, when the mole flow rate of methanol is 0.07 mol/s and the inlet temperatures are 373.2 K, 433.2 K, 473.2 K and 423.2 K, the methanol conversions are 0.83, 0.89, 0.93 and 0.97, respectively. However, when mole flow rate of methanol is smaller than 0.07 mol/s, the methanol conversion rearches almost 100% and the reaction of methanol steam reforming is nearly completed. The methanol conversion decreases with the increase of the methanol feeding rate at a fixed inlet temperature. The methanol conversion is affected by the reaction rate and the residence time of the reactant gases in the pore, which will increase with the increase of the reaction rate and the residence time. According to Eq. (6), the reaction rate increases with the increase of the inlet temperature of reactants, leading to the

Similar to the conventional methanol steam reforming, the mole ratio of water to methanol will influence the hydrogen production. Fig. 13 presents the variation of the methanol conversion with the solar radiation and the mole ratio of water/methanol. The

Fig. 12. Variation of outlet temperature with the methanol feeding rate and the inlet temperature.

280

Y. Wang et al. / Solar Energy 134 (2016) 273–283

feeding rate of reactants is fixed at 12.6 kg/h and the inlet temperature of the reactants is 473.2 K. Fig. 14 presents the influence of the solar radiation and the mole ratio of the water/methanol on the outlet temperature of the solar receiver/reactor. In Fig. 13, we can observe that the methanol conversion is a rising function with the mole ratio of the water-to-methanol for a fixed solar radiation intensity. With the increase of the mole ratio of water/methanol, the reaction, CH3OH + H2O ? CO2 + 3H2, can be achieved. The methanol conversion is a rising function of the solar radiation for a fixed mole ratio of the water-to-methanol when the DNI is smaller than 900 W/m2. One of possible reasons is that more solar flux may bring more solar thermal energy absorbed by the reactor and higher reactor temperature, resulting in a higher methanol conversion. When the DNI becomes larger than 900 W/m2, the methanol conversion ratio decreases a little with the increment of the solar radiation intensity, which is derived from the deactivation of the catalyst due to the high temperature. In Fig. 14, we present the influence rules of the solar radiation and the mole ratio of the water/methanol on the outlet temperature of the solar receiver/reactor. We can observe that the outlet temperature increases with the augmentation of the solar radiation at a fixed mole ratio of water/methanol, which is derived from the fact that more solar flux can lead to more solar thermal energy absorbed by the reactor and higher reactor temperature. The outlet temperature is a rising function with the mole ratio of the waterto-methanol for a fixed solar radiation intensity. The feeding rate of reactants is fixed at 12.6 kg/h, with the increase of mole ratio of water/methanol, the feeding rate of methanol becomes smaller, and thus the reacted methanol rate decreases, less thermal energy is absorbed by the steam methanol reforming reaction, and the excessive thermal energy heated the species.

3.4. A new solar receiver/reactor We find that owing to the characteristics of the chemical reaction of the methanol steam reforming, the required reaction heat is large at the beginning part of the chemical reaction in real-world applications. When the reaction is almost reacted, the required reaction heat becomes very small. In order to improve the thermochemical efficiency, the collected solar energy should match with the required reaction heat of the methanol steam reforming along the flow direction. The level of the concentrated solar thermal energy depends mainly on the shape of the concentrator and the

Fig. 13. Variation of methanol conversion with the solar radiation and mole ratio of water/methanol.

Fig. 14. Variation of outlet temperature with the solar radiation and the mole ratio of water/methanol.

receiver/reactor tube. In order to make the collected solar energy match the characteristic of the methanol steam reforming reaction, the collected solar energy at the inlet part of the receiver/reactor should be improved, while at the outlet part should be reduced. This can be implemented by improving the aperture width of the inlet part, meanwhile, reducing the aperture width of the outlet part, make the collected solar energy distributes non-uniformly along the flow direction. Thus, a new solar receiver/reactor is proposed by changing the aperture width along the flow direction. More solar energy is collected at the beginning part of the chemical reaction which improves the reaction rate of the methanol steam reforming. While smaller solar energy is collected at the latter part of the chemical reaction which reduces the solar energy that is absorbed by the gaseous species. Thus, more solar energy is used to drive the methanol steam reforming to improve the thermochemical efficiency of the mid-and-low temperature solar receiver/reactor. The advantages of such design can be outlined as follows: (1) Matching of energy levels and quantities between solar thermal energy and chemical reaction. For this design of a solar receiver/reactor, the employed endothermic reaction requires the concentrated solar energy to provide enough thermal energy for driving the reaction; at the same time, the grade of the provided solar thermal energy should be in good agreement with that of the reaction. The level of the solar thermal energy is dependent on the shape of the concentrator. The solar thermal energy provided by the new structural solar collector can be satisfied with the specific reaction characteristic of this methanol steam reforming. (2) Agreement of the receiver/reactor configuration with the dimension of the focal plane. A well-designed solar receiver/reactor has a characteristic structure for achieving both the desired chemical reactivity and the suitable concentrated solar energy. The configuration determines the kinetics of the chemical reaction, which relates to the conversion of the chemical reactant, and influences solar thermal energy to be converted into chemical energy. On the other hand, the distribution of the concentrated solar energy at the focal plane over the receiver surface also affects the chemical reactivity. For this new structural solar collector, a good matching of the reactor configuration to the focal

281

Y. Wang et al. / Solar Energy 134 (2016) 273–283

gso-ch

Q X CH3 OH nCH3 OH Dr H ; ¼ so-ch ¼ IA Q solar

ð15Þ

where Q so-ch is the energy that the solar thermal energy converted into the chemical energy, which is equal to the reaction enthalpy; Dr H stands for the enthalpy of the reaction corresponding to per mole of the reactant at a temperature of solar reactor and a pressure p in J mol
1. It is worth mentioning that the chemical energy of the species does not include the sensible heat. Fig. 17 illustrates the influence of the non-uniform solar flux distribution on the thermochemical efficiency. The thermochemical efficiency of the non-uniform cases are much larger than that of the uniform case. It can be observed from Fig. 17 that the thermochemical efficiency increases with the increase of the nonuniformity of the collected solar flux distributions along the flow

(a) Traditional solar collector

Conditions

Heat flux distribution (W/m2) x = 0–2 m

Heat flux distribution (W/m2) x = 2–4 m

Case Case Case Case

q0 1.25 q0 1.5 q0 1.75 q0

q0 0.75 q0 0.5 q0 0.25 q0

I II III IV

40

Solar flux/kW/m2

In order to investigate the influences of heat flux distributions, the total collected solar energy is kept constant. To make the calculation simple, the stepped change of the width of the parabolic trough collector is chosen, which is shown in Fig. 15. In this section, numerical simulations with five different conditions are implemented to investigate the influence of solar energy distributions along the flow direction on the performances of the mid-and-low temperature solar receiver/reactor. The five conditions of the collected solar energy distribution along flow direction are shown in Table 4, in which q0 is the solar flux distribution when the DNI is 600 W/m2, and it is shown in Fig. 16. Case I is a typical working condition of the solar receiver/reactor, in which the collected solar energy distribution is unchanged along the flow direction. For Cases I–IV, the collected solar energy decreases gradually along flow direction, while the total collected solar energy remains the same. The other conditions remain the same. The DNI is 600 W/m2, the mole flow rate of methanol is 0.06 mol/s, the inlet temperature of the reactants is 473.2 K and the mole ratio of water/methanol is fixed at 1:1 and the length of the receiver/reactor tube is 4 m. To further evaluate the thermochemical performances of the mid-and-low temperature solar receiver/reactor, thermochemical efficiency of gso-ch Liu et al., 2009, defined as the portion of solar energy converted into chemical energy, is given as:

Table 4 Collected solar flux distributions along flow direction under different conditions.

30

20

10

0

0

45

90

135

180

225

270

315

360

Circumferential angle/degree Fig. 16. Solar flux distribution on the outer surface of the receiver/reactor.

0.53

Thermochemical efficiency

plane, can improve the methanol conversion, the efficiency of solar energy converted into chemical energy, and avoid the deactivation of the catalyst bed.

Thermochemical efficiency

0.52 0.51 0.50 0.49 0.48

1

2

3

4

Working conditions Fig. 17. Thermochemical efficiency under different collected solar energy distributions.

(b) New structural solar collector

Fig. 15. Comparison of the traditional solar collector and new structural solar collector.

282

Y. Wang et al. / Solar Energy 134 (2016) 273–283

direction. The thermochemical efficiencies are 0.49 and 0.52 for the typical working condition and condition VI, respectively. The thermochemical efficiency is increased by three percentage points from Conditions I to VI. Fig. 18 illustrates the influence of the solar flux distribution along the flow direction on the maximum temperature of the catalyst bed. It can be seen that with the increase of the nonuniformity of the solar flux along the flow direction, the maximum temperature of the catalyst bed appears from the outlet part to the middle part of the solar receiver/reactor. The maximum temperature of the catalyst bed decreases with the increase of the nonuniformity of the collected solar flux distributions along the flow direction at first, when reaches Case II, the maximum temperature of the catalyst bed increases with the increase of the nonuniformity of the collected solar flux distributions along the flow direction, which is yet in a safety range. The performance of the receiver/reactor can be improved by improving the collected solar flux at the beginning part of the tube while reducing the collected solar flux at the outlet part.

(a) Temperature distributions of Case I

(b) Temperature distributions of Case II

4. Conclusions Based on the fluid flow and energy conservation equations coupling radiation/convection/conduction heat transfer with the reaction kinetics, a three-dimensional model is developed to numerically reveal the performances of the receiver/reactor for the solar hydrogen production with the methanol steam reforming. The main research findings can be summarized as follows.

(c) Temperature distributions of Case III

Maximum temperature of catalyst bed/ oC

(d) Temperature distributions of Case VI 275

Maximum temperature / oC

265

255

245

235

1

2

3

4

Working conditions

(e) Maximum temperature of the catalyst bed Fig. 18. Temperature distributions of the catalyst bed under different solar energy distributions.

(1) The factors influencing the hydrogen production and the catalyst bed temperature distributions, including the inlet methanol feeding rate, the solar radiation and the inlet temperature, are investigated. The methanol conversion ratio increases with the augmentation of the inlet temperature and decreases with the increase of the methanol feeding rate. When the inlet temperature is high or the inlet methanol feeding rate is low, the deactivation of the catalyst bed may appear. (2) The methanol conversion ratio increases with the increment of the mole ratio of the water-to-methanol and DNI. However, when the DNI is high, the deactivation may appear. For these cases simulated in this paper, when the DNI is larger than 900 W/m2, the methanol conversion ratio decreases a little with the increment of the DNI owing to the fact that a high temperature may lead to the deactivation of the catalyst. (3) A new solar receiver/reactor is proposed by increasing the collected solar flux at the beginning part of the tube while reducing the collected solar flux at the outlet part to match the level of the concentrated solar energy with the chemical reaction. Compared with traditional solar receivers/reactors, the thermochemical efficiency of the new structure solar receiver/reactor can be increased by 3%. For the proposed new solar receiver/reactor, the employed endothermic reaction requires the concentrated solar energy to provide enough thermal energy for driving the reaction; at the same time, the grade of the provided solar thermal energy should be in a good agreement with that of the reaction. In real-world applications, the understanding of the complicated physical and chemical mechanism of the hydrogen production by the steam methanol reforming method is beneficial for the efficient utilization of mid-and-low temperature solar thermal

Y. Wang et al. / Solar Energy 134 (2016) 273–283

energy. Our work provides an alternative approach for investigating the performances of the solar receiver/reactor and improving the thermochemical efficiency of the solar receiver/reactor, which needs to be further validated by practical applications in the future and to be further studied on the respects such as the kinetic model of the methanol steam reforming, the evaluation method, and the experimental test. Acknowledgements The authors appreciate financial support provided by the National Natural Science Foundation of China (Nos. 51276214, 51236008), the Project of Outstanding Young Scientists, Chinese Academy of Sciences. References Abanades, S., Flamant, G., 2006a. Thermochemical hydrogen production from a twostep solar-driven water-splitting cycle based on cerium oxides. Sol. Energy 80, 1611–1623. Abanades, S., Flamant, G., 2006b. Solar hydrogen production from the thermal splitting of methane in a high temperature solar chemical reactor. Sol. Energy 80, 1321–1332. Abuiadala, A., Dincer, I., 2012. A review on biomass-based hydrogen production and potential applications. Int. J. Energy Res. 36, 415–455. Chueh, W.C., Falter, C., Abbot, M., Scipio, D., Furler, P., Haile, S.M., Steinfeld, A., 2010. High-flux solar-driven thermochemical dissociation of CO2 and H2O using nonstoichiometric ceria. Science 330, 1797–1801. CN-31 light oil pre-reforming catalyst, 2012. Product Technical Data. Southwest Research & Design Institute of Chemical Industry. Vol. 9. Dufour, J., Serrano, D.P., Galvez, J.L., Moreno, J., Garcia, C., 2009. Life cycle assessment of processes for hydrogen production: environmental feasibility and reduction of greenhouse gases emissions. Int. J. Hydrogen Energy 34, 1370– 1376. Fogler, H.S., 1986. Elements of Chemical Reactor Engineering. Prentice-Hall, Englewood cliffs, NJ. Forristall, R., 2003. Heat transfer analysis and modeling of a parabolic trough solar receiver implemented in engineering equation solver. Technical Report, NREL/ TP-550-34169, October 2003. Gadalla, M.A., Ratlamwala, T.A.H., Dincer, I., 2010. Energy and exergy analyses of an integrated fuel cell and absorption cooling system. Int. J. Exergy 7, 731–754. Hirsch, D., Steinfeld, A., 2004. Solar hydrogen production by thermal decomposition of natural gas using a vortex-flow reactor. Int. J. Hydrogen Energy 29, 47–55. Hong, H., Jin, H.G., Ji, J., Wang, Z.F., Cai, R.X., 2005. Solar thermal power cycle with integration of methanol decomposition and middle-temperature solar thermal energy. Sol. Energy 78, 49–58. Hong, H., Liu, Q., Jin, H., 2009. Solar hydrogen production integrating low-grade solar thermal energy and methanol steam reforming. Trans. ASME J. Energy Res. Technol. 131, 012601-1. Hou, Z., Zheng, D., Jin, H., Sui, J., 2007. Performance analysis of non-isothermal solar reactors for methanol decomposition. Sol. Energy 81, 415–423.

283

Jiang, C.J., Trimm, D.L., Wainwright, M.S., Cant, N.W., 1993. Kinetic mechanism for the reaction between methanol and water over a Cu–ZnO–Al2O3 catalyst. Appl. Catal. A 97, 145–158. Jin, H., Sui, J., Hong, H., Wang, Z., Zheng, D., Hou, Z., 2007. Prototype of solar receiver/ reactor with parabolic troughs. Trans. ASME J. Sol. Energy Eng. 129, 378–381. Kanouff, M.P., Gharagozloo, P.E., Salloum, M., Shugard, A.D., 2013. A multiphysics numerical model of oxidation and decomposition in a uranium hydride bed. Chem. Eng. Sci. 93, 212–225. Krishna, R., 1990. Multicomponent surface diffusion of adsorbed species: a description based on the generalized Maxwell–Stefan equations. Chem. Eng. Sci. 45 (7), 1779–1791. Krishna, R., van Baten, J.M., 2005. Diffusion of alkane mixtures in zeolites: validating the Maxwell–Stefan formulation using M D simulations. J. Phys. Chem. B 109 (13), 6386–6396. Krishna, R., Wesselingh, J.A., 1997. The Maxwell–Stefan approach to mass transfer. Chem. Eng. Sci. 52 (6), 861–911. Kuhn, J., Stemmer, R., Kapteijn, F., Kjelstrup, S., Gross, J., 2009. A non-equilibrium thermodynamics approach to model mass and heat transport for water pervaporation through a zeolite membrane. J. Membr. Sci. 330, 388–398. Lanchi, M., Varsano, F., Brunetti, B., Murmura, M.A., Annesini, M.C., Turchetti, L., Grena, R., 2013. Thermal characterization of a cavity receiver for hydrogen production by thermochemical cycles operating at moderate temperatures. Sol. Energy 92, 256–268. Liu, Q., Hong, H., Yuan, J., Jin, H., Cai, R., 2009. Experimental investigation of hydrogen production integrating methanol steam reforming with middletemperature solar thermal energy. Appl. Energy 86, 155–162. Liu, Q., Jin, H., Hong, H., Sui, J., Ji, J., Dang, J., 2010. Performance analysis of a midand low-temperature solar receiver/reactor for hydrogen production with methanol steam reforming. Int. J. Energy Res. 35, 52–60. Muradov, N.Z., Veziroglu, T.N., 2008. Green path from fossil-based to hydrogen economy: an overview of carbon-neutral technologies. Int. J. Hydrogen Energy 33, 6804–6839. Poling, B.E., Prausnitz, J.M., John Paul, O.C., 2001. The Properties of Gases and Liquids, fifth ed. McGraw-Hill, New York. Price, H., Lüpfert, E., Kearney, D., Zarza, E., Cohen, G., Gee, R., 2002. Advances in parabolic trough solar power technology. J. Sol. Energy Eng. Trans. ASME 109, 109–125. Segal, A., Epstein, M., 2003. Optimized working temperatures of a solar central receiver. Sol. Energy 75, 503–510. Silva, C.M., Lito, P.F., 2007. Application of the Maxwell–Stefan approach to ion exchange in microporous materials batch process modelling. Chem. Eng. Sci. 62, 6939–6946. Steinfeld, A., 2005. Solar thermochemical production of hydrogen – a review. Sol. Energy 78 (5), 603–615. Steinfeld, A., Meier, A., 2004. Solar fuels and materials. Encycl. Energy 5, 623–637. Wang, Y.J., Liu, Q.B., Lei, J., Jin, H.G., 2015. Performance analysis of a parabolic trough solar collector with non-uniform solar flux conditions. Int. J. Heat Mass Transf., 236–249 Wendelin, T., 2003. SolTRACE: a new optical modeling tool for concentrating solar optics. In: Proceedings of ISEC 2003: International Solar Energy Conference, Hawaii, USA. ISEC2003-44090. Xu, B., Li, P.W., Chan, C.L., 2015. Energy storage start-up strategies for concentrated solar power plants with a dual-media thermal storage system. Trans. ASME J. Sol. Energy Eng. 137 (5), 051002. Zamfirescu, C., Dincer, I., 2014. Assessment of a new integrated solar energy system for hydrogen production. Sol. Energy 107, 700–713.