Reaction Characteristics of Ni-Based Catalyst Supported by Al₂O₃ in a Fluidized Bed for CO₂ Methanation

Abstract

CO₂ methanation is a promising technology to store renewable energy by converting carbon dioxide with green hydrogen into methane, which is known as power to gas (PtG). In this study, CO₂ methanation performance of a Ni/Al₂O₃ catalyst was investigated in a bubbling fluidized bed (BFB) and the axial gas concentration, temperature, and CO₂ conversion were densely analyzed. Moreover, a modified reaction kinetic model was proposed, and the results were compared with experimental data. The bed temperature increased by 11 °C from 340 °C to 351 °C within the first 30 mm of the fluidized bed. The CO₂ conversion was approximately 90% within 50 mm from the bottom of the reactor and was maintained above this height. The Ni/Al₂O₃ catalyst exhibited the highest CO₂ conversion (95%) at 320 °C. Using a simple plug-flow reactor model, three optimized kinetic modification factors (1.5094, 0.0238, and 0.2466) were used to fit the experimental data. The hydrodynamic effects significantly influenced the chemical reaction kinetics of the BFB.

1. Introduction

Over the last decade, the demand for sustainable and energy-efficient systems has increased considerably. Many countries are increasing the share of renewable energy in their total energy use, toward targeting carbon neutrality by 2050. A total of 181 GW of renewable power was added in 2018 at a pace consistent with that of 2017. The number of countries integrating high shares of variable renewable energy (VRE), such as solar and wind power generation, continues to rise [1]. However, the amount of curtailment (unused power) is rapidly increasing because of difficulties in power transmission and distribution facilities, intermittency of renewable energy sources, and unbalanced demand and supply of renewable energy [2,3,4]. Therefore, energy storage systems (ESS) that can store curtailed electric energy and stabilize the power supply are required to increase the share of renewables in the global energy generation. In an ESS, unused power is converted to other forms of energy and stored, from which electrical energy can be extracted when required. Some ESS technologies include pumped hydroelectric storage (PHS), compressed air energy storage (CAES), supercapacitor and battery energy storage, thermal energy storage, and power-to-gas (P2G; power-to-H₂ and power-to-CH₄) [5,6,7,8]. Among these, P2G is the most suitable for the long-term and high-capacity storage of renewable energy [5,9,10,11]. In particular, CO₂ methanation is a promising approach for storing renewable energy by converting carbon dioxide (CO₂) to methane (CH₄) using green hydrogen; this has an advantage in terms of the levelized cost of energy analysis (LCOE). Currently, most demonstration plants performing the CO₂ methanation process are operating mainly in Europe; these plants function either by catalytic or biological methanation. Catalytic methanation has a higher reaction rate and throughput than biological methanation [12,13].

CO₂ methanation:

$${\mathrm{CO}}_2+4{\mathrm{H}}_2 \rightleftharpoons {\mathrm{CH}}_4+2{\mathrm{H}}_2\mathrm{O} \to ∆\mathrm{H}=-168 {\mathrm{kJ} \mathrm{mol}}^{-1}$$

(1)

However, for catalytic methanation, the catalyst must be thermally stable to sustain the hot spot caused by the strongly exothermic methanation reaction. Further, the heat of the reaction needs to be dissipated to enhance the conversion efficiency [14,15,16]. Therefore, the removal of reaction heat is a critical issue in CO₂ methanation to avoid catalyst deactivation due to sintering and carbon deposition on the catalyst [17,18,19,20]. From a heat recovery point of view, a fluidized bed reactor with a high heat transfer capability is preferred to a fixed-bed reactor for exothermic reactions. So far, studies on CO₂ methanation using a fluidized bed reactor have focused on the identification of an appropriate catalyst [21,22,23,24,25,26], characterization of fluidization [27,28,29], efficiency of heat exchange [30], and modeling of a fluidized bed reactor [28,31,32]. Xiao et al. [33] studied the effect of the pyroelectric materials for reduction of CO₂ to chemical, and reported that the pyroelectric materials show excellent efficiency and great potential for CO₂ reduction utilizing environmental heat energy. A NiO-based catalyst can be employed as a pyroelectric catalyst. In the present study, the reduced form of a catalyst was used; thus, it is considered that the catalyst does not play a role in pyroelectric material. Le et al. [23] reported that a high Ni dispersion and large surface area of the support are crucial factors for CO₂ methanation. Italiano et al. [25] studied the effect of supports on catalytic deactivation, wherein the basic sites of the support had a positive effect on the catalytic activity, and the Ni-support interaction influenced the anti-coking and anti-sintering abilities. Unlike a fixed-bed reactor, the attrition of the catalyst and fluidization characterization are important factors in a fluidized bed reactor. This study aims (i) to fabricate durable Ni-based catalyst particles suitable for CO₂ methanation in bubbling fluidized bed (BFB) reactors, (ii) to optimize the process operation targeting maximum CO₂ conversion and CH₄ production rate, and (iii) to identify a flexible chemical reaction kinetic model for CO₂ methanation in BFB considering hydrodynamics effects. Herein, we prepared spherical Ni-based catalysts via a spray-drying method, and investigated the influence of temperature, pressure, feed gas ratio, and gas velocity conditions on the methanation performance of the as-prepared catalyst in a fluidized bed reactor. In addition, the gas concentration profiles were calculated using an intrinsic kinetic model and the results were compared with the experimental data. Based on this, a reaction kinetic model with modification factors was proposed for BFB.

2. Results

2.1. Characterization of Catalysts

Figure 1 shows the particle size distribution (PSD) and an image of the GMC-Y3 catalyst. The particle size of the catalyst and bulk density is ~100 μm and 1177 kg/m³, respectively. The GMC-Y3 catalyst was well formed into a spherical shape for use in a fluidized bed reactor. To determine the reducibility of NiO in the GMC-Y3 catalyst, H₂-temperature programmed reduction (TPR) results were obtained for the Ni-based catalysts calcined in air at 800 °C (Figure 2). The reduction of NiO in the GMC-Y3 catalyst starts at ~300 °C and ends at ~750 °C. In this study, the reduction temperature was set at 600 °C and maintained for 2 h.

To confirm the composition and phase of the catalysts, X-ray diffraction (XRD) and inductively coupled plasma spectrometry (ICP) analyses were conducted. Figure 3 shows the XRD patterns of the GMC-Y3 catalysts. The catalyst is a composite of the active material and support. The active material is NiO, as indicated by the diffraction peaks at approximately 37°, 43°, and 62°, which are assigned to the (111), (200), and (220) planes of cubic NiO, respectively [34]. The support materials used for resisting attrition were Al₂O₃, TiO₂, and NiTiO₃. Figure 3b shows the XRD pattern of the reduced GMC-Y3 catalyst at 600 °C; most of the NiO peaks disappear and new Ni peaks are identified at 44°, 53°, and 78°.

Figure 4 shows the N₂ adsorption-desorption isotherm and pore size distribution of the GMC-Y3 catalyst. The N₂ adsorption-desorption profile of the GMC-Y3 catalyst exhibits a typical type-II behavior characteristic of non-porous materials [35]. The GMC-Y3 catalyst has a small surface area (

Table 1. Textural properties and H₂ consumption of the GMC-Y3 catalyst.

Catalyst	SBet (m2/g)	Pore Volume (cm3/g)	Average Pore Diameter (nm)	Crystallite Size of NiO (nm) 1	H2 Consumption (mmol/gNi)	Dispersion of Ni (%)
GMC-Y3	4.7	0.012	12.25	90	0.00386	0.1

¹ Calculated by the Scherrer equation using the D/Max-2500 diffraction and Jade 8 program (SmartLab High temp/Rigaku, Japan).

). The surface area, pore volume, and average pore diameter are 4.7 m²/g, 0.012 cm³/g, and 12.5 nm, respectively. The specific surface area of the catalyst decreases due to the presence of promoters and binders, and also due to the high calcination temperatures used for hardening the catalyst; thus, catalyst attrition is an important factor affecting catalyst efficiency in fluidized bed reactors. The particle size of NiO is computed as 90 nm using the Scherrer equation, based on the NiO (111) diffraction peak obtained via XRD measurements.

2.2. Reactivity of Catalyst

The methanation performance of the GMC-Y3 catalyst in terms of the CO₂ conversion was evaluated under varying conditions of temperature, H₂/CO₂ ratio, pressure, and gas velocity, as shown in

Table 2. Experimental conditions for CO₂ methanation.

	Temperature (°C)	Pressure (Bar)	H2/CO2 Ratio	Uo/Umf
Case 1	260–380	1	4	2.5
Case 2	340	1–9	4	2.5
Case 3	340	1	3.0–5.0	2.5
Case 4	340	1	4	1.5–5.5

. Figure 5 shows the effects of multiple parameters on CO₂ conversion of the GMC-Y3 and the commercial catalyst (GMC-C) along with the equilibrium curve calculated by HSC Chemistry software 5.1 ((Metso Outotec, Finland). All experimental results corresponded to the equilibrium curves. The CO₂ conversion increased to 95% at 320 °C and then decreased at higher temperatures. Compared with the commercial catalyst, the conversion of GMC-Y3 was lower than that of GMC-C at a low temperature below 300 °C, while CO₂ conversion was almost similar to each other at a high temperature above 300 °C. Higher temperatures above 400 °C accelerate catalyst deactivation and decrease the selectivity toward CH₄, resulting in the production of CO gas. At a H₂/CO₂ ratio of 3, the CO₂ conversion is 75%, and the conversion increases as the H₂/CO₂ ratio is raised; the conversion reaches ~99% when the H₂/CO₂ ratio is 5. There is no notable difference in the conversion as a function of the reaction pressure from 1–9 bar, and the conversion in the entire range is ~90–92%. The effect of gas velocity on CO₂ conversion was confirmed at U₀/U_mf (ratio of the superficial gas velocity (U₀) over the minimum fluidization velocity (U_mf)) from 1.5 to 5.5. As shown in Figure 5d, the conversion decreases as U₀/U_mf increases. In a fluidized bed reactor, gas velocity is an important factor in CO₂ conversion. The number and size of bubbles increase as the gas velocity increases; thus, they can pass through the reactor without contact with the catalyst. This phenomenon reduces the residence time in the reactor and decreases the contact efficiency between the gas and catalyst. Thus, a higher gas velocity decreases CO₂ conversion (Figure 5d). However, the low gas velocity in the fluidized bed indicates that the throughput per unit volume of the reactor is low. Therefore, the optimal gas velocity should be determined by considering the CO₂ conversion and methane production rates. In conclusion, when comparing the activity performance of the GMC-Y3 catalyst and the commercial catalyst, GMC-Y3 is considered to be more suitable for the CO₂ methanation process.

Figure 6 shows the measured axial bed temperature, gas concentration, and conversion profiles at 1 bar, 340 °C, and 3.5 U_mf. The flow rates of CO₂, H₂, and N₂ gases were 90, 360, and 1550 mL/min, respectively. The feed gases (CO₂ and H₂) were diluted with N₂ to enable temperature control for the strongly exothermic reaction. As shown in Figure 6a, the temperature at the center of the catalyst increases from 340 °C to 351 °C owing to the exothermic reaction. The minimum height required to reach the maximum temperature is ~30 mm. The methanation reaction is rapid and almost terminates at the beginning of the reactor. The maximum temperature is maintained until the end of the fluidized bed. Above a height of 250 mm, the temperature decreases, which indicates that this is the freeboard region of the BFB.

2.3. Reaction Kinetics

The axial mole fractions (y_i) of H₂, CO, CH₄, and CO₂ in Figure 7 were used to identify the chemical reaction kinetics and estimate the three factors, viz. η₁, η₂, and η₃ in Equation (A1). ${\eta }_i$ represents the differences in catalyst characteristics to Xu and Froment [36] and hydrodynamics of BFB. Figure 7 compares the experimental data and kinetic model results for y_i. When the intrinsic kinetics model [36] without the three factors (η₁, η₂, and η₃) is used, the CO₂ methanation reactions occur quickly at the bottom of the bed and reach thermodynamic equilibrium (Figure 6a). However, the intrinsic kinetics model deviates considerably from the experimental data. In particular, at the top of the bed, the mole fraction (0.014) of H₂ obtained from the kinetic model is far from the experimental value (0.05). When the root mean square error (RMSE) is minimized using η₁, η₂, and η₃ (see Equation (3)), this deviation is reduced, and the reaction progresses slowly at the bottom (Figure 6b). The mole fraction of H₂ obtained from the modified kinetic model is 0.022 at the top of the bed (56% error between the experimental and model values). The reason for the larger error in the H₂ mole fraction than those of other components is that even a small difference in the reaction extent leads to a large deviation in the H₂ mole fraction, because four moles of H₂ participate in CO₂ methanation (see Equation (1)).

The optimized modification factors are 1.5094, 0.0238, and 0.2466, respectively, and the RMSE is 0.00118. The parity plot between the axial experimental data (y_i,exp.) and model prediction (y_i,model), including the error bound within ±30%, is shown in Figure 8. The mole fractions of CO, CO₂, and CH₄ are relatively well predicted, whereas those of H₂ deviate substantially from the experimental data. The kinetics model modified with η₁, η₂, and η₃ factors provides a similar tendency to that of the experiment in the BFB. However, a large discrepancy between the model and experimental results was observed at the bottom of the BFB, indicating a rapid temperature rise. The kinetic model does not adequately represent the hydrodynamics of solid particles, as mentioned in Section 3. A three-dimensional hydrodynamic model combined with reaction kinetics is appropriate for the BFB reaction. Nevertheless, the simple plug-flow model is a valuable tool for the rapid fitting of kinetic parameters with experimental data. Therefore, a hybrid kinetic model that considers hydrodynamic effects is suitable for BFB.

3. Materials and Methods

3.1. Preparation of CO₂ Methanation Catalyst

Three types of catalysts for CO₂ methanation were prepared; GMC-Y3 was identified as the optimal catalyst for the CO₂ methanation reaction in the fluidized bed reactor. Thus, the subsequent experiments and analyses were performed with GMC-Y3 (see Appendix A), prepared using a spray dryer. The catalyst was composed of the active material (50% NiO, Daejung Chemicals & Metals, 99% Republic of Korea) and supports (98% Al₂O₃, 98% TiO₂, and 25~32% CaSiO3·nH₂O, Daejung Chemicals & Metals, Republic of Korea). The GMC-Y3 catalyst was mixed with 50 wt % NiO, 20 wt % TiO₂, 20 wt % Al₂O₃, and 10 wt % CaSiO₃ and then milled for 12 h. The prepared slurry was formed into spherical particles by spray-drying. The temperature of the air supplied to the spray dryer is about 260 °C, and the temperature of the spray dryer chamber is about 120 °C. In addition, the rotation speed of the atomizer (disc type) was fixed to about 6000 rpm. The dried catalysts were calcined at 800 °C in air for 4 h. The calcined catalyst was sieved to control the particle size from 106 to 212 μm.

3.2. Fluidized Bed Reactor for CO₂ Methanation

Figure 9 shows the lab-scale fluidized bed reactor used in this study. The main reactor was made of stainless steel (SUS 310) with an inner diameter of 0.05 m, thickness of 0.003 m, and height of 0.68 m. The temperature of the reactor was controlled based on the internal temperature of the reactor (K-type thermocouple), using an electric furnace with a temperature indicator. The temperature was measured at a height of 0.05 m from the bottom of the reactor. In addition, a K-type thermocouple with an inner diameter of 0.25 inch was inserted into the tube, and the tube was set to move vertically by the stepping motor. The differential pressure of the fluidized bed was measured by installing a differential pressure transducer and pressure indicator at a height of 0.01 m and 0.58 m, respectively, from the reactor bottom. The reactant gases were introduced into the reactor using a mass flow controller (MFC), and the outlet gases from the reactor were automatically analyzed in real time using an online gas analyzer (Advance Optima, AO2020, ABB, Schwaebisch Hall, Germany). The thermodynamic equilibrium data for CO₂ methanation were calculated using the HSC chemistry software.

3.3. Experimental Conditions and Procedure

The GMC-Y3 catalyst was loaded into the fluidized bed reactor at a height of 0.25 m, and the temperature was increased to a reduction temperature of 600 °C while maintaining a flow of N₂ gas (0.076 m/s). When the reactor temperature reached 600 °C, H₂ gas (20 vol %) was introduced. At this H₂ concentration, the reduction is regarded as complete. Subsequently, the temperature was cooled to 300 °C while purging N₂ gas. Next, CO₂ and H₂ gases were introduced into the reactor to initiate CO₂ methanation. We have investigated the influence of four conditions: temperature, pressure, CO₂/H₂ ratio, and gas flow rate on CO₂ conversion (Table 2). The mole fraction of CO₂ conversion, $X_{\mathrm{C}{\mathrm{O}}_2}$, is defined as follows:

$$X_{{\mathrm{CO}}_2}=\frac{C_{{\mathrm{CH}}_4}+C_{\mathrm{CO}}}{C_{{\mathrm{CH}}_4}+C_{{\mathrm{CO}}_2}+C_{\mathrm{CO}}}$$

(2)

4. Reaction Kinetics Model

The intrinsic reaction kinetics model proposed by Xu and Froment (1989) [36] for Ni-based catalysts supported by Al₂O₃ was used in this kinetic study. The detailed kinetic model is presented in Appendix B. Because the bubbling fluidized bed (BFB) shows complex particle hydrodynamics, and also axially and radially distributed temperatures, the intrinsic model derived under ideal conditions must be modified for the BFB. Three values of η for the three reactions were applied in Equation (A1) to consider the effects of catalyst material characteristics, hydrodynamics of BFB, diffusion limit inside the catalyst pellets, and uneven temperature, on the global reaction kinetics. The values of η were estimated by minimizing the RMSE of the CO₂, H₂, and CH₄ mole fractions between the kinetic model and experimental data using the trust-region reflection algorithm [37].

$$\underset{{\eta }_1,{\eta }_2,{\eta }_3}{\min }\mathrm{RMSE}=\sqrt{\frac{1}{N}{{\displaystyle \sum }}_{i=1}^N{\left(y_{i,exp.}-y_{i,model}\right)}^2},$$

(3)

where N is the number of experimental data points (=104) and $y_{i,exp.}$ and $y_{i,model}$ are the species mole fractions obtained from the experimental and kinetic models, respectively. The three modification factors were bound within 0 < η < 2 in this nonlinear optimization problem.

5. Conclusions

CO₂ methanation is a promising technology to store the intermittent renewable energies via the power-to-gas (PtG) process. In the present study, CO₂ methanation performance of the Ni/Al₂O₃ catalyst was evaluated in a bubbling fluidized bed (BFB) reactor to assess the effects of temperature, pressure, H₂/CO₂ ratio, and gas velocity on the methanation reactivity. A modified reaction kinetic model is also proposed, and the results were compared with experimental data. The temperature and gas concentration inside the fluidized bed reactor was investigated using a stepping motor, and the main reaction occurred within 50 mm of the bottom of the reactor. In the lower part of the reactor, the conversion was 90%, and the temperature increased by 11 °C from, 340 °C to 351 °C within the first 30 mm of the fluidized bed. The CO₂ conversion was approximately 90% within 50 mm from the bottom of the reactor and was maintained above this height. The optimum temperature for methanation using the GMC-Y3 catalyst was 320 °C to achieve the highest CO₂ conversion of 95%. Three kinetic modification factors, 1.5094, 0.0238, and 0.2466, were optimized to fit the experimental data using a simple plug-flow reactor model and the common intrinsic kinetic rate equations for Ni-based catalysts. Deviations between the fitted kinetic model and experimental data are attributed to hydrodynamic effects. Thus, a hybrid kinetic model that considers the hydrodynamic effects of BFB is necessary.