Numerical Modeling of Solar Thermochemical Reactor for Kinetic Analysis Academic research paper on "Chemical engineering"

CrossMark

Available online at www.sciencedirect.com

ScienceDirect

Energy Procedía 49 (2014) 735 - 742

SolarPACES 2013

Numerical modeling of solar thermochemical reactor for kinetic

analysis

Selvan Bellan, Elisa Alonso, Carlos Perez-Rabago, José Gonzalez-Aguilar*,

Manuel Romero

IMDEA Energy Institute, Ramon de la Sagra 3, 28935 Móstoles, Spain

Abstract

A lab-scale thermochemical reactor is designed and fabricated for the solar-driven thermal reduction of non-volatile manganese oxide to produce hydrogen by water splitting thermo chemical cycles. A time dependent three dimensional numerical model is developed to investigate the performance of the reactor since the chemical kinetics strongly depends on irradiance, temperature and fluid flow distribution around the reactant. Radiation heat transfer is calculated by using surface-to-surface (S2S) radiation model. Thermo-fluid flow, absorption efficiency and the temperature distribution of the sample are predicted as a function of time and the model is validated by experimental measurements.

© 2013 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.Org/licenses/by-nc-nd/3.0/).

Selectionandpeerreview bythescientificconference committeeofSolarPACES2013underresponsibilityofPSEAG.

Final manuscript published as received without editorial corrections.

Keywords: Chemical reactor; CFD modeling; Concentrated solar energy; Numerical analysis;

1. Introduction

Solar fuel produced by concentrated solar energy through endothermic chemical reaction is one of the most attractive research areas in solar thermochemical processes [1]. As the high temperature is demanded for achieving high conversion rates, concentrated solar energy has been used as heat source e.g. [2]. Since the production of fuels by thermochemical reactors considered as an effective method, a specific effort has been given to design and optimize different kind of thermochemical reactors [3]. To attain and withstand high reaction temperatures with

* Corresponding author. Tel.: +34 91 737 11 20; fax: +34 91 737 11 40. E-mail address: jose.gonzalez@imdea.org

1876-6102 © 2013 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/Kcenses/by-nc-nd/3.0/).

Selection and peer review by the scientific conference committee of SolarPACES 2013 under responsibility of PSE AG. Final manuscript published as received without editorial corrections. doi: 10.1016/j.egypro.2014.03.079

sufficient efficiency, cavity-type receivers made up of ceramic materials have been used, whereby the absorption of the receiver is increased by focusing concentrated solar radiation through a small aperture e.g. [4].

In the past few decades, significant progress has been achieved in the development of optical systems for large scale solar concentration and capable to reach more than 2000 K, which are required for the efficient two step thermochemical cycles [5]. The first step is the dissociation of metal oxide to metal or lower valance metal oxide by endothermic reaction using solar energy. The second step is the hydrolysis of metal to produce hydrogen and the corresponding metal oxide by non-solar exothermic reaction. One of the most favorable metal oxide pairs for two step thermochemical cycle is apparently ZnO/Zn. Thermal dissociation of this redox pair has been extensively studied e.g. [6] . A three-step thermochemical cycles have also been investigated for hydrogen production such as Mn2O3/MnO redox pairs e.g. [7]. Recently, thermal reduction of non-volatile metal oxides by directly exposing the reactant to high radiation fluxes have been investigated by our group using lab-scale thermochemical reactor [8, 9], and the instantaneous oxygen concentration at the outlet is measured using ZrO2-cell oxygen analyzer. The main objective of this investigation is to develop a transient 3D numerical model to study the thermal performance of the reactor since the chemical reactions involved in the thermochemical cycles as described in aforementioned processes depend on the heat transfer process involved in the reactant and its atmosphere.

2. Model description

The cross-sectional view of the computational domain is shown in Fig. 1, which is formed by the region limited by quartz glass, the inlet region, ceramic cylinder surrounded by the cavity and the reactor exit. Concentrated radiation enters into the cavity receiver through transparent quartz glass window and aperture, and impinges over the front surface of sample, which is mounted on a rod type sample holder and placed inside the cavity receiver as shown in fig. 1. To reduce the conduction heat losses, the cavity receiver is protected by well insulating layer. Carrier gas is injected radially through four inlet ports close to the quartz window. A 7-kWe solar simulator is used as radiation source. The complete description of this reactor can be found in [8, 9, 10]. To predict the instantaneous temperature distribution and fluid flow inside the reactor, a numerical model is developed by assuming the flow is laminar and the reactor cavity walls are opaque and lambertian reflectance. So, the governing equations to simulate the thermo-fluid flow inside the reactor are given by

SAMPLE

ALUMINA

Fig.1. Computational domain used for simulation

dp ~dt"

■V.(p u)= 0

p— + p(u.V)u = -VP + V dt

^(vu + (Vu)T )-| ^Vu

FCP — + PCpu.VT = V(kVT )

where p, u, P, n, k, T, and Cp are density, velocity vector, pressure, dynamic viscosity, thermal conductivity, temperature, and specific heat respectively. In solid regions, the convection term in eq. (3) is neglected. Radiation heat transfer is calculated by using surface-to-surface (S2S) radiation model [11] and the boundary conditions are given in table 1.

Table 1. Boundary conditions

Boundary

(CD&PO)

Irradiated walls (GH,AB,EF,MN,QR)

Solid-fluid interface (FG,HI, IJ, JK,LM)

Quartz glass (AR)

External boundary (AC, DK, LO, PR)

Outlet

Temperature

q = e(Gs - aT4) + qd q = e(G- oTA) + qa q = eo(T4mb - TA) q = KTamb - T) ST/Sn=0

Velocity V;

Mass flow rate (kg/s)

dVJdn=0

In table 1, q, s, a, G, h, Tamb and qd are heat flux, emissivity, Stefan-Boltzmann constant, incident radiation flux, heat transfer coefficient, ambient temperature and diffusive heat flux between the solid-fluid interface respectively.

Inlet: The carrier gas tube from the gas cylinder is divided into four equal parts and radially connected to the reactor near to the quartz window at perpendicular to each other. The argon gas is supposed to enter into the cavity (CD and PQ in Fig.1) at 1 atm and 300 K, where the mass flow rate of 6.69 x 10-5 kg/s is applied at each inlet. Thus the total mass flow rate 0.00027 kg/s is given to the reactor, which is equal to 9 LPM (litre per minute).

Irradiated walls and Solid-fluid interface: Since the thermochemical reactor horizontally fixed opposite to the solar simulator, it is assumed that the concentrated radiation directly impinging over the frustum surface (AB, EF, MN, and QR) and front surface of the sample (GH) through quartz window. Hence the net heat flux boundary condition is applied at these surfaces as given in table 1. Where the first term represent the incident radiation flux and the second term represent the diffusive heat flux due to the carrier gas flow over the surfaces, where Gs is calculated by

Gs = Go0lar + G = Go0lar +\Fss,J<dS (4)

Where Gsolar is the incident radiation coming from the solar simulator, G is the incident radiation from the different emitting surfaces, which is calculated by the radiosity (J') and view factor (Fss), which is the fraction of energy leaving surface s' that is incident on surface s. The radiosity at every surface point inside the cavity receiver

Frustum radius (m)

0.02 0.03 0.04 0.05

\ I —i—i— I ' ' — I ' — 'I

0 0.003 0.006 0.009 0.012

Sample radius (m)

Fig.2. Incident radiation flux distribution along the sample (GH) and frustum (AB, EF, MN, QR) surfaces.

is calculated by surface to surface radiation model. The radiation flux distributions at these surfaces (Gsoiar) are measured experimentally and shown in fig. 2. To measure the concentrated radiation distribution on the targeted surface, the commonly used techniques of direct and indirect methods are applied. The direct method is based on the single point flux measurement using a radiometer [12], in this study which is done with a gardon type calorimeter (Vatell TG1000). The indirect method is based on the measurement of the flux distribution of flat lambertian target from the image analysis of CCD camera [13-14]. In this investigation, gardon-type calorimeter measurements are used to convert the gray-scale images of CCD camera to get the flux distribution. Thus, the radiation flux distribution of the target surface is obtained by combining the both techniques measurements. The error bars are obtained by the standard deviation of the measurements using the calorimeter, the maximum of 1.69% is found. Chemical reactor is placed collinear with solar simulator optical axis and the radiation flux distribution at the target surface is measured. This radiation flux is given as boundary condition.

In order to ensure the overall energy balance and the diffusive heat flux continuity across the solid-fluid interface, a simplified method to deal the boundary layer treatment, namely the harmonic mean of thermal conductivities have been employed immediately adjacent to the solid surface (about 1 mm thickness) to evaluate the diffusive heat flux at the interface [15]. Then the diffusive heat flux across the fluid-solid interface is given by,

where ks, kf, Ts and Tf are thermal conductivities and temperatures of the solid and fluid respectively.

Quartz glass and External boundary: Since the ambient surroundings behave as blackbody, the surface to ambient radiation boundary condition is applied and the heat flux condition is applied at the external boundary (AC, DK, LO and PR) of cavity cylinder with the heat exchange coefficient of 5.0 W/m2 K.

The thermo physical properties of argon and reactor components are given in [10]. Governing equations are solved using the finite element technique based commercial software Comsol Multiphysics 4.3 [11]. Grid independence tests are conducted on the mesh models at the beginning of the simulation process. The mesh is refined in near wall and high (temperature and velocity) gradient regions. In this study three levels of refinements are considered (195584, 344409 and 558641 number of elements) and from these tests it is concluded that the mesh having 195584 number of element is adequate since it is grid independent.

' 2kskf ^ Tf - Ts

ks + kf x2 - x,

V s J y 2 1

3. Results and discussion

Since our main objective of this investigation is to analyse the thermal performance of the reactor, a graphite rod is placed in the sample holder instead of the reactant since it is a good reference material due to its well-known thermal and optical properties. With this arrangement, the sample is directly exposed to the concentrated radiation. It is also assumed that the concentrated radiation from the solar simulator is not fluctuating. At time t = 0 the initial temperature throughout the reactor is assumed to be 300 K, when t > 0 the boundary conditions given in table 1 are applied.

The instantaneous temperature distribution of the reactor at vertical plane (ZX) is shown in fig 3. Temperature distribution at the centreline of the sample and at the inner cavity wall of the reactor is shown in fig.4 (a) and (b) respectively as a function of time. Radiation coming from the solar simulator enters the chemical reactor through the quartz window and impinges on the frustum and the sample front region. As the sample is placed close to the focal point of the reactor, the incident radiation flux at the front surface of the sample is high and consequently this region initially absorbs the radiation and starts to heat up and transfer the heat throughout the sample rapidly. Since these walls are assumed as opaque-gray-diffuse surfaces, a portion of the incident radiation is reflected and diffused throughout the cavity and gradually heats up the inner cavity surface by radiation heat transfer along with the convective heat transfer due to the carrier gas flow along the cavity and sample walls. It is also observed that, the conduction heat transfer takes place between the sample and cavity through the sample holder since it is attached to the bottom side of the cavity, consequently temperature at the bottom side is higher than the top side of the cavity receiver.

In order to monitor the experiments, four thermocouples (T1-T4) are embedded inside the reactor at different positions as shown in fig 1. While conducting the experiments, temperature at those positions are measured as a function of time. The numerically predicted and experimentally measured temperatures at those positions are compared in Fig. 5. The predicted results are comparable with measurements except at T3. This may be due to the heat exchange coefficient at the external boundary (AC, DK, LO and PR), which could be lower than the assumed value.

Time= 10(e) Time=60(s) Time=120(s)

0 50 100 150 200 0 50 100 150 200 0 50 100 150 200

^ i'^-j ili'oo i4'oo is'oo li'oo

Fig.3. Temperature distribution of the reactor (at vertical plane-ZX) as a function of time.

Fig. 4. Centreline (a), and inner cavity wall (b) temperature profiles as a function of time.

a) 1200

ro 1000

a. F 800

*-*- F * * *

- i* __e>__0 — O- _Q_ _0 _ £ __Û--0 — O--

o & 0 o <> v - v -

k « « ■ ■ ■ ■ ■

6000 Time (s)

-" T1 sim

--T2 sim

T3 sim

----T4 sim

* T1 exp T2 exp « T3 exp T4 exp

Fig.5. Predicted and measured temperature at different positions (T1-T4) inside the reactor as a function of time.

To calculate the solar energy absorption efficiency of the reactor, the 1st law from the thermodynamic analysis of solar thermochemical processes, described by [5], is used. It is defined as the ratio of the net rate at which energy is being absorbed (Qabs) to the solar power coming from the concentrator (Qsoiar),

absorption

\irad V Qsolar J

Where Qrad = Qsolar - Qabs is the reactor radiation heat losses. The total incident radiation at sample front surface from solar simulator is 746.11 W. The net absorbed energy, net radiation losses and absorption efficiency, maximum temperature attained by the sample are predicted as a function of time and shown in fig 6(a) and (b) respectively. As expected, the net radiation absorption is high at initial stage; it absorbs almost the total energy except the reflected radiation. Then, the net absorption energy is gradually decreasing when the temperature of the sample increased. The net absorption power is reduced mostly by radiation losses when the temperature of the sample is above 1000 K [5]. Thus, the absorption efficiency and the maximum temperature attained by the sample are respectively

Fig.6. (a) Net absorbed energy and radiation losses; (b) absorption efficiency and maximum temperature attained by the sample as a function of

decreasing and increasing rapidly until 1000 K, and this effect is gradually decreasing with increasing the time as shown in Fig. 6.

4. Summary and conclusion

A time dependent three dimensional numerical model is developed to investigate the lab-scale thermochemical reactor for kinetic analysis and validated by using experimental measurements. Thermo-fluid flow inside the reactor, absorption efficiency, radiation loss and maximum temperature attained by the sample are predicted as a function of time. From this study it is concluded that, this model can be used to predict the solar reactor performance, optimise the reactor design and to implement the reaction kinetics.

Acknowledgements

Comunidad de Madrid and Structural Funds are acknowledged for its financial support to the SOLGEMAC project through the Programme of Activities between Research Groups (S2009/ENE-1617). JGA acknowledges support from the Spanish Ministry of Science and Innovation (grant Ramon y Cajal RYC-2009-05358).

References

[1] Romero M, Steinfeld A. Concentrating solar thermal power and thermochemical fuels. Energy Environ. Sci.2012. 5, 9234

[2] Kodama T, Gokon N. Thermochemical cycles for high temperature solar hydrogen production. Chemical Reviews. 2007. 107, 4048-4077.

[3] Meier A, Ganz J, Steinfeld A. Modeling of a novel high-temperature solar chemical reactor. Chemical Engineering Science. 1996. 51 (11), 3181-3186.

[4] Steinfeld A, Schubnell M. Optimum aperture size and operating temperature of a solar cavity-receiver. Solar Energy. 1993.50(1),19-25.

[5] Steinfeld A, Palumbo R. Solar thermochemical process technology. In: Meyers, R.A. (Ed.), Encyclopedia of Physical Science and Technology, 15. Academic Press. 2001. 237-256.

[6] Schunk L.O, Haeberling P, Wepf S, Wuillemin D, Meier A, Steinfeld A. A receiver-reactor for the solar thermal dissociation of zinc oxide. Journal of Solar Energy Engineering. 2008. 13 0(2), 021009-15.

[7] Sturzenegger M, Nuesch P. Efficiency analysis for a manganeseoxide-based thermochemical cycle. Energy. 1999. 24 (11), 959-970.

[8] Alonso E, Gómez F, González-Aguilar J, Romero M. Experimental analysis of Mn3O4/MnO reduction in a packed-bed type solar reactor: Oxygen partial pressure influence. Proceedings of the SolarPACES Conference, Granada, Spain. 2011. 20-23.

[9] Alonso E, Perez-Rabago C, Gonzalez-Aguilar J, Romero, M. Thermal performance and residence time distribution determination in a solar reactor for chemical kinetics, 18th SolarPACES Conference, Marrakech. Morocco. 2012.

[10] Bellan S, Alonso E, Gomez-Garcia Fabrisio G, Perez-Rabago C, Gonzalez-Aguilar J, Romero M Thermal performance of lab-scale solar reactor designed for kinetics analysis at high radiation fluxes. Chemical Engineering Science. 2013. 101, 81-89.

[11] COMSOL Multiphysics Version 4.2, COMSOL AB, Stockholm, Sweden, 2012

[12] Gardon R. An instrument for the direct measurement of intense thermal radiation, Review Sci. Instruments, 1953. 24 (5), 366-370.

[13] Neumann, A., Monterreal, R. Measurement of concentrated solar radiation with the HERMES II systema at PSA, 6th International Symposium on Solar Thermal Concentrating Technologies, Mojacar, Spain, Sept. 28- Oct. 2, 1992.

[14] Gómez, F., Gonzalez-Aguilar, J., Romero, M.. Experimental 3D flux distribution of a 7 kWe-solar simulator. Proceedings of the SolarPACES Conference, Granada, Spain, 20-23, 2011.

[15] Patankar S.V. Numerical Heat Transfer and Fluid Flow. Hemisphere/McGraw-Hill, Washington, DC (Chapter 3), 1980.