Experimental Study of Cement-Supported CuO Oxygen Carriers in

Feb 1, 2013 - The Reduction Kinetic of the Combined Cu-Based Oxygen Carrier Used for Chemical Looping Gasification Technology. Kun Wang , Weipeng Luan...
5 downloads 10 Views 4MB Size
Article pubs.acs.org/EF

Experimental Study of Cement-Supported CuO Oxygen Carriers in Chemical Looping with Oxygen Uncoupling (CLOU) Lei Xu, Jianan Wang, Zhenshan Li,* and Ningsheng Cai Key Laboratory for Thermal Science and Power Engineering of Ministry of Education, Beijing Municipal Key Laboratory for CO2 Utilization & Reduction, Department of Thermal Engineering, Tsinghua University, Beijing 100084, People’s Republic of China ABSTRACT: Chemical looping with oxygen uncoupling (CLOU) provides a possible process that could be used for combustion of solid fuels with easy separation of CO2. Cu-based oxygen carriers are among the most promising metal oxide oxygen carriers for CLOU, because CuO has a very high oxygen transport capability and high reaction rates. However, Cu-based oxygen carriers always suffer from the problem of agglomeration, which would result in defluidization. Inhibiting the agglomeration using methods as simple as possible becomes a key issue for the use of Cu-based oxygen carriers in CLOU. This study prepared four types of oxygen carriers with different CuO contents as the active component and cement as the support material by mechanical mixing. The oxygen release and uptake rates were determined by successive cycling tests in a TGA. The fluidization behavior against agglomeration during cyclic oxygen release/uptake reactions were tested in a batch-scale fluidized bed. Experimental results showed that the addition of cement can effectively inhibit the agglomeration of Cu-based oxygen carriers and this type of oxygen carrier could have potential use in the CLOU process. A theoretical equation based on the concept of Zener pinning force was developed to understand the effect of various support materials on the agglomeration limit.

1. INTRODUCTION The CO2 concentration in the atmosphere keeps rising in recent decades as the combustion of fossil fuels increases, which causes global warming. Carbon capture and storage (CCS) refers to technologies that could mitigate the large emissions of CO2 into the atmosphere from fossil fuel combustion in the power generation and other industrials by capturing CO2 and storing CO2 in a large inventory such as geologic formations. Chemical looping combustion (CLC), first proposed by Richter,1 is a type of CCS technology that allows the use of fossil fuels with inherent CO2 separation at a low energy cost and becomes a very promising combustion technology for the power generation and industrials. A key issue in the application of CLC is to find suitable oxygen carriers which should have high reaction activities, high stabilities, good fluidization properties, and low economic costs, as well as be environmentally friendly. The first research relating to oxygen carriers can be found in the study of reduction and oxidation cyclic characteristics of α-Fe2O3 medium proposed by Nakano in 1986.2 The majority studies in CLC are aimed at solid fuels and gas fuels. Because the solid− solid reaction rate is basically slower than that of the gas−solid reaction, the solid fuels for CLC are proposed to be used in three ways: (1) the solid fuels are first gasified in a gasifier and then the product gas (mainly H2 and CO) are introduced into the fuel reactor to react with oxygen carriers; (2) the gasification agent and solid fuels are introduced into the fuel reactor simultaneously which implies that both the gasification and the combustion occur in the same reactor; and (3) chemical looping with oxygen uncoupling (CLOU), which was first proposed by Mattisson et al.3 in 2005. In the CLOU process, the solid fuels react with the gas-phase oxygen released by the oxygen carriers. Thus, the combustion efficiency of solid fuels can be enhanced as the low reactivity of coal gasification is avoided. Meanwhile, the total amount of oxygen carriers needed in the reaction system and the reactor size can be reduced.3 © 2013 American Chemical Society

Oxygen carriers for the CLOU process should be capable of releasing and uptaking suitable amount of O2 in the temperature range for CLOU (typically 800−1200 °C). Thus, three metal oxide systems have been investigated as oxygen carriers for the CLOU process: CuO/Cu2O, Mn2O3/Mn3O4, and Co3O4/ CoO.3 The oxygen capability of Mn-based oxygen carriers is relatively low and the Co-based oxygen carriers are thermodynamically unfavorable and expensive. Cu-based oxygen carriers have high reaction rates, high oxygen capabilities, and low costs, which should be most promising for the CLOU process. However, the application of Cu-based oxygen carriers is mainly limited by the agglomeration problem. The ICB-CSIC group and Chalmers group conducted a series of studies on Cu-based oxygen carriers,4−8,11,12 which found that the support materials such as α-Al2O3, γ-Al2O3, ZrO2, MgAl2O4, and NiAl2O4−Al2O3 can inhibit the agglomeration. The study by Wen et al.9 showed that the stability decreased as the cyclic reaction proceeds and the addition of CoO, NiO, MgO, and SrO2 can increase the stability. Natural copper ores have been investigated as potential oxygen carriers for the CLOU process for its low cost, but also suffer from the agglomeration problem.10 Many types of oxygen carriers have been made and investigated for use in the CLOU process, as summarized in Table 1.10−17 As can be seen in Table 1, many support materials were tested for the development of copper-based oxygen carriers in CLOU or CLC. Among these support materials, the use of Al2O3 is favored, because of its abundance. However, the formation of CuAl2O4 in Al2O3-supported CuO oxygen carriers will result in partial loss of CuO, which is not favorable for the CLOU process. Some of other support materials are also expensive, and the Received: December 2, 2012 Revised: January 31, 2013 Published: February 1, 2013 1522

dx.doi.org/10.1021/ef301969k | Energy Fuels 2013, 27, 1522−1530

Energy & Fuels

Article

Table 1. Oxygen Carriers in the Literature for CLOU metal oxide (wt %)

support material (wt %)

preparation methoda

size (μm)

sintering temperature and time

facilityb

CuO (40) CuO (60) MnO2 (68.3)

ZrO2 Al2O3 MgO

FG FG FG

180−250 125−180 125−180

950 °C, 6 h 1300 °C, 4 h 1100/1125/1150/1200/1300 °C

bFB bFB bFB

MnO2 (66.7)

MgO/Ca(OH)2 (28.2/5.1) MgO/Ca(OH)2 (31.9/7.9) MgO

FG

125−180

1100/1150/1200 °C

bFB

FG

125−180

1100 °C

bFB

FG

125−180

1100/1200/1300 °C

bFB

γ-Al2O3 MgAl2O4 Al2O3 TiO2 SiO2 sepiolite ZrO2 MgAl2O4 sepiolite ZrO2 SiO2

WI WI PE PE PE PE PE PP PP PP PP SD crushing

100−300 100−300 100−300 100−300 100−300 100−300 100−300 100−300 100−300 100−300 100−300 125−180 200−500

850 °C, 1 h 850 °C, 1 h 950/1100 °C, 6 h 950/1100 °C, 6 h 950/1100 °C, 6 h 950/1100 °C, 6 h 950/1100 °C, 6 h 950/1100/1300 °C, 6 h 950/1100/1300 °C, 6 h 950/1100/1300 °C, 6 h 950/1100/1300 °C, 6 h 1100 °C 950 °C, 3 h

TGA TGA TGA, bFB TGA, bFB TGA, bFB TGA, bFB TGA, bFB TGA, bFB TGA, bFB TGA, bFB TGA, bFB bFB TGA, bFB

MnO2 (60.2) Mn3O4 (65.4) CuO (15/33) CuO (15/33) CuO (60/80)

CuO (60/80)

Mn2O3/Fe2O3 copper ore

reacting atmosphere

ref(s)

coke, air CH4, air CH4, N2, O2(10%) + N2 CH4, N2, O2(10%) + N2 CH4, N2, O2(10%) + N2 CH4, N2, O2(10%) + N2 N2, CO2, air N2, CO2, air N2, CO2, air N2, CO2, air N2, CO2, air N2, CO2, air N2, CO2, air N2, CO2, air N2, CO2, air N2, CO2, air N2, CO2, air coke, air N2, CO2, air

11, 12 13, 14 15 15 15 15 16 16 16 16 16 16 16 16 16 16 16 17 10

a

FG = freeze granulation; WI = wet impregnation; PE = pelletizing by extrusion; PP = pelletizing by pressure; SD = spray drying. bbFB = batch fluidized bed, TGA = thermogravimetric analyzer.

search and test of inexpensive support material for copper-based oxygen carrier is still necessary and meaningful for the development of CLOU. Therefore, in this work, the very inexpensive cement was proposed to be used as the support material, and cement-supported Cu-based oxygen carriers with different CuO contents and different calcination temperatures were tested in a TGA and a batch-scale fluidized-bed reactor, with a focus on the issues of reactivity and agglomeration. Results showed that the addition of cement can effectively inhibit the agglomeration of copper-based oxygen carriers and moreover, the preparation method is simple and the raw materials are inexpensive.

Figure 1. Photo of the hardened samples. in a thermogravimetric analyzer (TGA) (TA Instruments, Model SDT Q600). The reaction temperature and sample weight were continuously recorded by a data acquisition system connected to the TGA. A small amount of each sample (∼10 mg) was placed in a quartz sample holder and brought from the ambient temperature to 950 °C under pure CO2 flow for decomposition. After stabilization, multicycle tests were conducted continuously at alternating oxygen uptake and release conditions. The oxygen uptake temperature was set at 800 °C and the atmosphere was air with a flow rate of 100 mL/min. The oxygen release temperature was set at 950 °C and the atmosphere was pure CO2 with a flow rate of 100 mL/min. The rate of temperature change between 800 and 950 °C was set at 10 °C/min. 2.3. Experiments in the Fluidized-Bed Reactor. Oxygen uptake and release cycles were also carried out in a batch-scale fluidized bed, as shown in Figure 3. The experimental setup consisted of two quartz tubes: an inner tube and an outer one. The outer tube (30 mm in diameter) was heated by an electric furnace with quartz sand as the bed material and N2 as the fluidizing gas. The inner tube (19 mm in diameter) was packed with oxygen carrier particles (6.4 g). The inner tube can be lifted out of the outer tube so that the fluidization state can be observed during the test. The flow rates of introduced gases were controlled by mass flow controllers. The oxygen release test was conducted in pure CO2 as the fluidizing gas and the oxygen uptake test was conducted with air as the fluidizing gas. The flow rates of fluidizing gases for both oxygen uptake and release were 0.7 L/min (standard

2. EXPERIMENTAL DESCRIPTION 2.1. Preparation of the Cement-Supported CuO Oxygen Carrier Particles. Four different oxygen carriers were prepared in this investigation by mechanical mixing methods. First, powders of copper oxide and cement were uniformly mixed with different CuO contents (45, 35, 25 wt %). Then, water was added to the mixed powders and then the obtained slurries were dried in air at room temperature for 10 h, during which time the cement hardening process occurred. Afterward, the hardened samples were calcined in a muffle oven for 3 h at 950 °C (CuO content: 45, 35, 25 wt %) or 1150 °C (CuO content: 35 wt %). The four oxygen carriers are referenced in this study as 25%-950, 35%950, 45%-950, and 35%-1150. The hardened samples are shown in Figure 1. Finally, the hardened samples were crushed into particles and sieved to the size range of 200−500 μm. The phase composition was analyzed semiquantitatively by XRD. The main phase for the four oxygen carriers are copper oxide and calcium aluminate, as shown in Figure 2. Cuprous oxide was detected in the 35%-1150 sample, which is mainly due to the decomposition of CuO in the high-temperature calcination step. It is noteworthy from the XRD analysis that the interaction of copper oxide with cement was not found and CuO phase remained unchanged. 2.2. Oxygen Uptake and Release on Cement-Supported CuO Oxygen Carriers with TGA. Continuous cyclic tests of the oxygen uptake and release processes of the four oxygen carriers were carried out 1523

dx.doi.org/10.1021/ef301969k | Energy Fuels 2013, 27, 1522−1530

Energy & Fuels

Article

Figure 2. XRD analysis of (a) sample 35%-950 and (b) sample 35%-1150. monitoring the O2 content in the product gas. The oxygen uptake period was ∼5 min and the oxygen release period was ∼15 min. Twenty (20) cycles of oxygen uptake and release for each sample were conducted by the same procedure in the same system. Complete oxygen release was achieved in all cycles for all tests. 2.4. Characterization Techniques. The phase composition was detected by X-ray diffraction (XRD, Model D8, Discover). The microstructure of oxygen carrier was investigated by scanning electron microscopy (SEM, Model JSM-7001F, JEOL) coupled with an energydispersive X-ray (EDX) analysis system.

3. RESULTS 3.1. Oxygen Carrier Reactivity in TGA. In this study, the reactivity of the four different oxygen carriers was tested in TGA under well-defined conditions. The conversion (X) was defined as X=

MCu 2O mt − m0 0.5MO2 m0xCu 2O

(1)

where MCu2O and MO2 are the molecular weight of Cu2O and O2, respectively, mt is the sample mass at time t, m0 is the initial sample mass, and xCu2O is the Cu2O mass fraction. Theoretically, 160 g pure CuO should release 16 g of O2 corresponding to a maximum O2 release capacity of 0.10 g-O2/g-CuO at 100% CuO decomposition. Twenty (20) cycles of oxygen uptake and release were carried out for all four oxygen carriers. The three oxygen carriers calcined at 950 °C showed similar kinetics while the sample calcined at 1150 °C showed a different behavior in kinetics, as shown in Figure 4. The fifth cycle was chosen to make comparisons of the oxygen uptake and release rates between different samples. The

Figure 3. Experimental setup for multicycle tests in a batch-scale fluidized bed. state). The inlet gas was preheated as it flowed upward into the preheating region of the reactor. The heated gas flowed upward through the oxygen carrier particles and exited from the top of the reactor. The outlet gas was first filtered and then introduced into the gas analyzer. The process of the oxygen uptake and release was followed by

Figure 4. (a) Oxygen uptake and (b) release rate of the four oxygen carriers in the fifth cycle. 1524

dx.doi.org/10.1021/ef301969k | Energy Fuels 2013, 27, 1522−1530

Energy & Fuels

Article

Figure 5. (a) Oxygen uptake and (b) release rates of sample 35%-950 at different cycles.

Figure 6. (a) Oxygen uptake and (b) release rate of sample 35%-1150 at different cycles.

oxygen uptake rates of the fifth cycle for the three 950 °C calcined samples are basically the same, while the 1150 °C calcined sample shows the same in the initial stage (before 50 s) but becomes much slower after 50 s. The oxygen release rates are slightly different for the three 950 °C calcined samples and show a tendency that, as the CuO content increased, the rate became slower. The reaction rate of sample 35%-1150 is roughly the same as that of sample 45%-950 in the initial stage. Because the time for oxygen uptake in the TGA test is not sufficient for the oxygen carrier to fully oxidize, due to the slow oxygen uptake rate after the initial stage of each cycle, the final conversion of oxygen release failed to reach 100% for sample 35%-1150, as shown in Figure 4. The oxygen uptake and release rates of different cycles for each sample were also investigated. The three samples calcined at 950 °C shows the similar results as to the conversion change over time. Figure 5 shows the oxygen uptake and release rate of the 950 °C calcined sample with 35% CuO content in cycles 1, 5, 20, and 21. The curves of cycles 1, 5, and 20 were obtained from samples within a continuous 20-cycle test in the TGA, while the curve of cycle 21 was obtained from samples after a 20-cycle test in the fluidized bed. The oxygen release rate of cycle 1 is slightly higher than cycles 2−20 while the rates from cycle 2 to cycle 20 remains nearly the same, as illustrated by cycle 5 and cycle 20 in Figure 5. It is noteworthy that cycle 21 exhibits higher reaction rates for both oxygen uptake and release than cycles 1−20. As the curve of cycle 21 was obtained from samples after 20-cycle test in a fluidized bed, the rate increase must be due to the different reaction conditions between the TGA and the fluidized bed. The SEM-EDX results shows that the content of CuO on the surface

of oxygen carriers after reaction in fluidized bed is higher than that before reactions, which may indicate that CuO have a tendency to diffuse to the surface during the redox cycles in a fluidized bed. Thus, longer residence time and more-severe reacting conditions in the fluidized bed may result in the higher reaction rate indicated by the curve of cycle 21. Similar results were also obtained from sample 35%-1150 after reacting 40 cycles in the fluidized bed, as shown in Figure 6. Figure 6 shows the oxygen release and uptake rates of sample 35%-1150 for different cycles. Both the conversion and the reactivity of oxygen carriers increases with the increasing cycle numbers, as can be clearly seen in Figures 6 and 7. The three oxygen carriers calcined at 950 °C show good stability during the multicycle test in TGA. The weight losses were ∼4.5%, 3.5%, and 2.5%, respectively, for samples 45%-950, 35%-950, and 25%-950, as shown in Figure 7. These values agree well with the theoretical oxygen release amount. However, the weight loss of sample 35%1150 is less than the theoretical value. The XRD analysis shows that the phase composition of sample 35%-950 and sample 35%1150 are nearly the same. It can also be inferred that no chemical interaction between CuO and cement during the preparation process and the CuO phase remains unchanged. But the oxygen uptake rate of sample 35%-1150 slows quickly after the initial stage (50−100 s, as shown in Figures 4 and 6), a full conversion cannot be reached during the limited test period. Thus, the capability of sample 35%-1150 is less than the theoretical value. This may be due to the fact that the high calcination temperature (1150 °C) reduced the surface area of the particle and, as the cyclic reaction went on, the particle experienced a slight gain in porosity, which leads to the reactivity. 1525

dx.doi.org/10.1021/ef301969k | Energy Fuels 2013, 27, 1522−1530

Energy & Fuels

Article

smaller particles below 200 μm. The attrition behavior during the fluidization is another important parameter for the selection of oxygen carriers. In this study, the attrition rate is evaluated as the mass percentage of particles