Article pubs.acs.org/EF
Oxygen Release and Oxidation Rates of MgAl2O4‑Supported CuO Oxygen Carrier for Chemical-Looping Combustion with Oxygen Uncoupling (CLOU) Mehdi Arjmand,†,* Martin Keller,† Henrik Leion,† Tobias Mattisson,‡ and Anders Lyngfelt‡ †
Department of Chemical and Biological Engineering, Division of Environmental Inorganic Chemistry, Chalmers University of Technology, SE-412 96 Göteborg, Sweden ‡ Department of Energy and Environment, Division of Energy Technology, Chalmers University of Technology, SE-412 96 Göteborg, Sweden ABSTRACT: The chemical-looping combustion (CLC) and chemical-looping with oxygen uncoupling (CLOU) processes are novel solutions for efficient combustion with inherent separation of carbon dioxide. These processes use a metal oxide as an oxygen carrier to transfer oxygen from an air reactor to a fuel reactor, where the fuel reacts with the solid oxygen carrier. When solid fuel is used in CLC, the char must be gasified by, e.g., steam to form H2 and CO, that can be subsequently oxidized to H2O and CO2 by the oxygen carrier. In the case of CLOU, the oxygen carrier releases oxygen gas in the fuel reactor. This enables a high rate of conversion of char from solid fuels, because it eliminates the need for the gasification step needed in normal CLC with solid fuels. In this work, the rate of oxygen release and oxidation of an oxygen carrier consisting of CuO supported by MgAl2O4 (40/60 wt %) for the CLOU process is investigated. The oxygen carrier was produced by freeze-granulation, calcined at 950 °C, and sieved to a size range of 125−180 μm. The reaction rates were obtained in a laboratory-scale fluidized-bed reactor in the temperature range of 850−900 °C, under alternating reducing and oxidizing conditions. The rate of oxygen release was obtained using devolatilized wood char as the fuel in N2 fluidization. Care was taken to obtain reliable measurements not affected by the availability of the fuel and temperature increase in the bed during combustion of the fuel with the released oxygen from the carrier. The Avrami−Erofeev mechanism was used to model the rates of oxygen release and the values of ko and Eapp were estimated to be 2.5 × 105 s−1 and 139.3 kJ mol−1, respectively. The rates of Cu2O oxidation were investigated in a flow of 5% O2 at the inlet of the reactor. However, it was observed that the oxidation rate is limited by the oxygen supply, indicating rapid conversion of the oxygen carrier. From the obtained reaction rates, the minimum total amount of the investigated oxygen carrier needed in the air and the fuel reactor is estimated to be between 69−139 kg MWth−1.
1. INTRODUCTION As suggested by the intergovernmental panel on climate control (IPCC), a 50%−85% reduction in total CO2 emission by 2050 is necessary to limit the anticipated global temperature rise to 2 °C.1 Several alternative technologies have been proposed to mitigate the rising levels of CO2 in the atmosphere. Among these, carbon capture and storage (CCS) is considered promising. The chemical-looping combustion (CLC) process allows intrinsic separation of pure CO2 from hydrocarbon combustion. In a CLC system, two reactors, a fuel reactor and an air reactor, are interconnected.2−4 When fuel and air are introduced into the respective reactor, the following reactions occur: (2n + m)MexOy + CnH 2m → (2n + m)MexOy − 1 + nCO2 + mH 2O
2MexOy − 1 + O2 → 2MexOy
(1) Figure 1. Schematic of the chemical-looping combustion process. (2)
Here, MexOy and MexOy−1 are the fully oxidized and reduced forms of the oxygen carrier. The scheme of the process is shown in Figure 1. In the case of complete conversion of the fuel, the exhaust stream from the fuel reactor consists of only CO2 and H2O, from which pure CO2 could be obtained after the condensation of water. The reduced form of the oxygen carrier, MexOy−1, is then transferred to the air reactor, where it is © 2012 American Chemical Society
reoxidized by air, making it ready for the next cycle. The oxidation reaction is always exothermic while the reduction Received: June 14, 2012 Revised: September 26, 2012 Published: October 3, 2012 6528
dx.doi.org/10.1021/ef3010064 | Energy Fuels 2012, 26, 6528−6539
Energy & Fuels
Article
continuous operation21−25 and thermogravimetric studies,13,14,26−31 with and without supports. Among various supports for CuO oxygen carriers, Al2 O3 has received considerable attention.12,15−17,21,23,24,27,30−33 In addition, CuO−Cu2O is one of the prominent oxide pairs suitable for the CLOU process.10,34 CuO decomposes to Cu2O as per equation 5 when the actual concentration of oxygen is lower than the equilibrium concentration.
reaction can be exothermic or endothermic, depending on the nature of the carrier and the fuel.5 However, the sum of the heat from reactions 1 and 2 is the same as that for conventional combustion. Considering the minor energy penalty for fluidization and transportation of particles between the air and the fuel reactor, the CLC process still obtains a lower energy penalty for CO2 separation, compared to other CO2 capturing processes. CLC has been successfully demonstrated in many units of sizes up to 120 kW.6 Overviews of current achievements in CLC are given by Lyngfelt,6,7 Hossain and de Lasa,8 and Adanez et al.9 The reactivity of the oxygen carrier during oxidation and reduction and the ability to fully convert the fuel are among the most sought-after criteria. In addition, thermal stability, mechanical strength, fluidizability, and resistance to attrition and agglomeration are important. In order to achieve this, the active phase (i.e., the reactive metal oxide) is often mixed with an inert support such as TiO2, SiO2, ZrO2, Al2O3, or MgAl2O4.8 The CLC process can be used with gaseous, liquid, or solid fuels. In the case of solid fuels, the char remaining after devolatilization is gasified in the presence of steam, producing CO and H2, which can then react with the oxygen carrier. However, in the case of chemical-looping with oxygen uncoupling (CLOU),10 the char reacts directly with gaseous oxygen released from the oxygen carrier. Thus, in comparison to CLC where the reduction of oxygen carrier and oxidation of the gaseous fuel generally occurs in a single step, an additional step is needed in CLOU for the release of gaseous oxygen from the carrier prior to conversion of the fuel according to MexOy → MexOy − 2 + O2
4CuO ↔ 2Cu 2O + O2
As a result, oxygen is released, thereby making CLOU possible. The equilibrium oxygen concentration for CuO oxygen carriers is between 1.5% and 4.5%, in the temperature range of 900−950 °C,10 which would make oxidation difficult to achieve in this temperature range, using 5% O2.34 The optimum temperature in the fuel reactor could be set higher, since this will increase the equilibrium oxygen partial pressure and rate of oxygen release; however, considering the very important restraint that the air reactor temperature must be low enough to enable the oxidation, the temperature in the fuel reactor will, in practice, be determined by the temperature of the air reactor and the heat balance. The latter will be controlled by the solids circulation rate, and the temperature and composition of the incoming fuel and fluidization gases, as well as the heat of reaction. Because the reaction in the fuel reactor is exothermic, the temperature there will be somewhat higher than the temperature in the air reactor. Thus, from a CLOU point of view for CuO oxygen carriers, the optimum temperatures of the air and fuel reactors are likely in the range of 850−900 °C to facilitate both oxidation and oxygen release. Some studies have also investigated the use of CuO oxygen carriers in CLC/CLOU for solid fuels.10,13,14,22,32−38 Despite these attractive features, use of CuO as an oxygen carrier is not without limitations. For example, copper oxide suffers from a tendency toward agglomeration upon full reduction,15 because of the formation of elemental copper, which has a rather low melting temperature (1085 °C). However, a higher temperature in the fuel and/or the air reactor(s) may be an advantage, with respect to kinetics. In order to achieve this, the CuO oxygen carrier must be able to resist agglomeration or defluidization at higher temperatures while providing stable reactivity. Recent investigations21,23 have expanded the frontier of high-temperature CLC application of CuO oxygen carriers to 900 °C in continuous operation, without considerable particle attrition or operational difficulties. In the case of Al2O3 as a support, another difficulty arises due to the facile interaction between CuO and Al2O3 either during synthesis or during operation, resulting in the partial loss of CuO via the formation of copper(II) aluminate (CuAl2O4) and copper(I) aluminate (CuAlO2, delafossite).17,21,23,39 However, since the copper aluminate phases are highly reducible,16,17,21,23,29,39,40 the interaction between the support and the active phase does not necessarily cause a problem, with respect to CLC application. For CLOU however, this interaction must be avoided in order to retain CuO as an active phase. Since the interaction of CuO with Al2O3 seems to be difficult to avoid, other supports such as TiO2, ZrO2, SiO2 or MgAl2O4 likely need to be employed for CLOU.13,14,17,21−23,28,29,34 The increase of carbon conversion rate in the vicinity of CuO particles was noted by Lewis et al.,41 and it was associated with the direct oxidation of char, as with CLOU. However, the rate of oxygen release could not be obtained, because of the interference of direct oxidation with the gasification of the fuel caused by the fluidization in CO2. Later, the rate of char conversion was
(3)
This is followed by the normal combustion of the fuel via ⎛ m⎞ CnH 2m + ⎜n + ⎟O2 → nCO2 + mH 2O ⎝ 2⎠
(5)
(4)
The reduced oxygen carrier is transferred to the air reactor for reoxidation. The net heat of reaction for the CLOU processes is the same as CLC, and only the mechanism by which oxygen is accessed by the fuel differs. However, when using solid fuels such as coal, the CLOU process avoids the slow gasification of the solid fuel needed to produce syngas as a prerequisite for the reaction with the oxygen carrier.10 The oxygen carrier in CLOU must be able to release O2 and oxidize at temperatures suitable for the process (i.e., 800−1200 °C). Conventional CFB boilers often operate at an air ratio of 1.2,2 which corresponds to an oxygen concentration of 5% at the outlet of the air reactor in a realistic CLC unit. Higher air ratios give additional cost for air supply and heat loss in flue gases. Therefore, it is preferable to have a maximum outlet concentration of 5%, which means that oxide systems with an equilibrium partial pressure low enough for oxidation (i.e., below 5% O2) are required. However, the oxide should be able to release a significant concentration of oxygen in the fuel reactor. Therefore, such thermodynamic and kinetic requirements limit the choice of oxygen carriers for the CLOU process. Oxides of transition metals (Mn, Fe, Co, Ni, and Cu), their mixtures, and several natural ores have been used as oxygen carriers in CLC.6−9 Copper oxide has received a great deal of attention as an efficient oxygen carrier, because of its high reactivity, high oxygen transport capacity, and absence of thermodynamic limitation for complete combustion of the fuel. Research has been conducted using copper oxide as an oxygen carrier in fluidized-bed11−18 and fixed-bed19,20 batch reactors, 6529
dx.doi.org/10.1021/ef3010064 | Energy Fuels 2012, 26, 6528−6539
Energy & Fuels
Article
reported having a 45-fold increase,34 compared to a synthesized MgAl2O4-supported Fe2O3 and ilmenite (a natural mineral) oxygen carriers, as determined using Mexican petroleum coke and a 40 wt % CuO/ZrO2 oxygen carrier. Chadda et al.42 obtained the rate of pure CuO decomposition in N2 in the temperature range of 760−910 °C using TGA. Eyring et al.38 simulated the conditions of CLOU in TGA, using pure CuO powder, and verified the results obtained previously. Sahir et al.43 also studied the decomposition kinetics of CuO and oxidation rates of Cu2O based on previously reported results of the CuO/ ZrO2 oxygen carrier using Mexican petroleum coke as fuel.34 Reduction and oxidation kinetics of CuO oxygen carriers for CLC has also been successfully acquired in TGA.44−46 Chuang et al.47 obtained the rate of oxidation of Cu and Cu2O for Al2O3supported CuO oxygen carrier in a fluidized-bed reactor within 300−750 °C. Subsequently, the rate of reduction of the same oxygen carrier was obtained in the temperature range of 450− 900 °C using H2 and CO in N2.40 More recently, Adánez-Rubio et al.48 investigated the CLOU operational regions for a CuObased oxygen carrier and showed that the solid fuel could be fully converted with very low solids inventory. The knowledge of the reduction and oxidation rates is of great importance in the design phase and for determination of the solids inventory in CLC/CLOU. In a previous investigation,17 a freeze-granulated CuO oxygen carrier supported by MgAl2O4 (40/60 wt %), showed stable chemical-looping properties during 17 cycles of redox using methane as fuel and 5% O2 in N2. Because of the absence of interaction between the active phase (CuO) and the support phase (MgAl2O4), the oxygen carrier showed intact content of the CuO after the final cycle. Similar results have also been reported by Adánez-Rubio et al.13 and Abad et al.22 using a spray-dried 60 wt % CuO supported by MgAl2O4 after screening of several CuO-based oxygen carriers.14 These results show the suitability of MgAl2O4-supported CuO oxygen carriers for the CLOU process. In this work, an attempt was made to obtain the oxygen release and oxidation rates of the freeze-granulated CuO supported by MgAl2O4 oxygen carrier17 for the CLOU process and the results are reported herein.
instantaneous freezing. The material is then freeze-dried at a temperature corresponding to the vapor pressure of water at −10 °C. This was followed by calcination at 950 °C for 6 h at a ramp rate of 5 °C/ min. The calcined material was sieved through stainless steel screens to yield particles in the range of 125−180 μm. 2.2. Characterization of the Oxygen Carrier. The oxygen carrier was analyzed before and after the experiments using powder X-ray diffraction (Siemens, D5000 Diffractometer) with Cu Kα radiation. The morphological investigation was carried out using an environmental scanning electron microscopy (ESEM) system fitted with a fieldemission gun (FEI, Quanta 200). The BET surface area of the CO2 particles was evaluated by N2 absorption (Micromeritics, TriStar 3000). The porosity of the particles was determined using mercury intrusion (Micromeritics, AutoPore IV 9500). The effective density of the particles, 125−180 μm in size, was measured, assuming a theoretical void fraction of 0.37 of a packed bed with uniform spherical particles. The crushing strength of the particles was measured as the strength needed to fracture a single particle sized within 180−250 μm. An average of 30 tests per sample was obtained using a digital force gauge (Shimpo, FGN-5). The crushing strength was found to be less than 0.5 N. However, as will be shown later in Section 3.6, the low crushing strength did not cause any problem, with respect to agglomeration or fragmentation for particles during the rate determination experiments. Nonetheless, for use in a full-scale plant, the crushing strength would need to be increased, e.g., by optimizing the manufacturing technique. 2.3. Experimental Setup and Procedure. To obtain the rate of release of oxygen from CuO, the oxygen carrier must be exposed to an oxygen-deficient environment at a suitable temperature to allow for the decomposition of CuO (see reaction 5). However, when nitrogen is used to simulate such a condition, reaction 5 may approach the equilibrium oxygen concentration, thus releasing oxygen at a limited rate, which may be slower than the actual rate in a real process, where the released oxygen will be consumed by a fuel. Thus, such an experimental approach would not readily yield the relevant rate of oxygen release since CuO decomposition is hindered by the equilibrium concentration of oxygen surrounding the particles. To overcome the equilibrium restriction, the experiments were carried out in a well-mixed fluidizedbed apparatus using a devolatilized Swedish wood char (Skogens Kol AB) as fuel. Thus, the fuel will consume the oxygen released, keeping the ambient O2 concentration low. Therefore, the rate of char combustion equals the rate of decomposition of CuO in the oxygen carrier. Table 2 shows the analysis of the wood char, as received and after devolatilization of the fuel. The wood char was crushed and sieved to
2. EXPERIMENTAL SECTION
Table 2. Analysis of the Swedish Wood Char Used as Fuel in This Work, As Received and after Devolatilization
2.1. Preparation and Manufacturing of the Oxygen Carrier. The oxygen carrier used in this investigation and its physical properties are summarized in Table 1. The particles were manufactured by freezegranulation. Here, a water-based slurry of CuO and support powder MgAl2O4 (S30CR, Baikowski) with weight ratio of 40/60 wt %, along with a small amount of dispersant (Dolapix PC21), is prepared. The mixture is then ball-milled for 24 h and a binder (poly(vinyl alcohol), PVA) is added prior to granulation. The slurry is pumped through a spray nozzle and into liquid nitrogen to form spherical particles upon
ultimate [wt %] (dry basis) as-received after devolatilization
N
C
H
S
9.7 14.3
0.51 0.71
78.7 88.9
2.9
