Investigation of Chemical Looping Combustion by ... - ACS Publications

Jul 11, 2006 - Energy Balance Analysis. The oxygen carrier reduction process by solid fuel is far more complicated compared to that by gaseous fuels...
5 downloads 0 Views 281KB Size
Download PDF
1836

Energy & Fuels 2006, 20, 1836-1844

Investigation of Chemical Looping Combustion by Solid Fuels. 1. Process Analysis Yan Cao and Wei-Ping Pan* Institute for Combustion Science and EnVironmental Technology, Western Kentucky UniVersity, Bowling Green, Kentucky 42101 ReceiVed July 24, 2005. ReVised Manuscript ReceiVed December 17, 2005

To concentrate CO2 in combustion processes by efficient and energy-saving ways is a first and very important step for its sequestration. Chemical looping combustion (CLC) could easily achieve this goal. However, only limited references are available that use coal as the energy resource in a CLC process even though the development of CLC of solid fuels follows the trend of energy utilization. This paper is the first in a series of two, where we present the concept of a CLC process of solid fuels using a circulating fluidized bed with three loop seals. The riser of this circulating fluidized bed was used as the oxidizer of the oxygen carrier; one of the loop seals was used as the reducer of the oxygen carrier and the separator for ash and oxygen carrier, and the other two loop seals were used for pressure balance in the solid recycle process. Pressure profiles of recycled solids using this process are presented in detail. For the development of an oxygen carrier, we focused on the establishment of a theoretical frame of oxygen transfer capability, reaction enthalpy, a chemical equilibrium, and kinetics. Analysis results indicated that Cu-, Ni-, and Co-based oxygen carriers may be the optimum oxygen carriers for the CLC of solid fuels. Mn-based oxygen carriers have several disadvantages in their lower oxygen transfer capability, thermodynamic limitations of purifying the CO2 stream, or a larger endothermic reduction enthalpy. Fe-based oxygen carriers have the disadvantage of a larger endothermic enthalpy in the reducer and lower reactivity. Thermodynamic analysis indicated that CO2 can be concentrated and purified to at least 99% purity for the gas-solid reaction mode (reduction of the oxygen carrier by gasification products such as CO and H2) or even higher for the solid-solid reaction mode (reduction of the oxygen carrier directly by solid fuels) on the basis of the selected oxygen carriers. A Cu-based oxygen carrier is the choice that has the potential to make the reducer self-sustaining or autothermal because of its exothermic nature during reduction. This would be beneficial for simplifying the operation of the reducer. The tendency of the Cu-based oxygen carriers to agglomerate can be eliminated by decreasing the operating temperature in the CLC system (600900 °C). In the second part of the series, we will evaluate the reduction kinetics of selected Cu-based oxygen carriers by coal and other “opportunity solid fuels” using a simultaneous differential scanning calorimetrythermogravimetric analysis to simulate a microreactor, using an X-ray diffractometer and a scanning electron microscope to characterize the solid residues, and a thermogravimetric analysis coupled with mass spectra to characterize the evolved gas compositions.

1. Introduction It has been known for more than 100 years that CO2 is a greenhouse gas.1 Fossil fuel combustion is a major source of CO2 emission. Worldwide fossil-fuel-based power production contributes about one-third of the total CO2 emissions annually.2 In a conventional combustion system, fuel is directly mixed with air and burned. This results in a low partial pressure of CO2 in the presence of a majority of nitrogen. To dispose of the diluted CO2 emissions will create a significant energy penalty.3 The estimated cost of the CO2 disposal, ∼$4-8 U.S. per ton of carbon,4 is much smaller when compared to the costs for the separation and purification of the diluted CO2 stream, which is typically in the range of $100-200 U.S. per ton of carbon.5 For a coal-fired power plant, roughly one-fifth of the electricity * To whom correspondence should be addressed. Fax: 270-745-2221. E-mail: [email protected]. (1) Arrhenius, S. On the Influence of Carbonic Acid in the Air upon the Temperature of the Ground. Philos. Mag. (1798-1977) 1896, 41, 237277. (2) Herzog, H.; Drake, E.; Adams, E. CO2 Capture, Reuse, and Storage Technologies for Mitigating Global Climate Change. A White Paper; U.S. Department of Energy: Washington, D.C., January, 1997, U.S.DOE/DEAF22-96PC01257. (3) Herzog, H.; Eliasson, B.; Kaarstad, O. Capturing Greenhouse Gases. Sci. Am. 2000, 282 (2), 72-79.

produced will be lost to CO2 separation and compression because of its low concentrations in flue gases.6 Faster CO2 sequestration by chemical measures at combustion sources with high CO2 concentrations (i.e., stationary power plants) would be highly desirable. Among available or proposed technologies7,8 involving CO2 purification in the combustion process, chemical looping combustion (CLC) is the most promising technology to combine (4) Yu, J.; Corripio, A. B.; Harrison, D. P.; Copeland, R. J. Analysis of the Sorbent Energy Transfer (SETS) for Power Generation and CO2 Capture. AdV. EnViron. Res. 2002. (5) Riemer, P. Greenhouse Gas Mitigation Technologies. An Overview of the CO2 Capture, Storage and Future Activities of the IEA Greenhouse Gas R& D Programme. The IEA Greenhouse Gas R&D Programme, 1998. http://www.ieagreen.org.uk/paper2.htm (accessed Dec 2005). (6) Freund, P. Abatement and Mitigation of Carbon Dioxide Emissions from Power Generation. Powergen 98 Conference, Milan, June, 1998. http:// www. ieagreen. org. uk/pge98. htm (accessed Dec 2005). (7) Lyngfelt, A.; Leckner, B. Technologies for CO2 Separation. In Minisymposium on CO2 Capture and Storage; Lyngfelt, A., Azar, C., Eds.; Publisher: Location, 1999; pp 25-35. Chalmers University of Technology and University of Gothenburg, Goteborg, Sweden, October 22, 1999. Available on http://www.entek.chalmers.se/∼anly/symp/sympco2.html (accessed Dec 2005). (8) Croiset, E.; Thambimuthu, K. Coal Combustion with Flue Gas Recirculation for CO2 Recovery. In Greenhouse Gas Control Technologies; Riemer, P., Eliasson, B., Wokauun, A., Eds.; Elsevier: New York, 1999; pp 581-586.

10.1021/ef050228d CCC: $33.50 © 2006 American Chemical Society Published on Web 07/11/2006

Chemical Looping Combustion by Solid Fuels

Energy & Fuels, Vol. 20, No. 5, 2006 1837

fuel combustion and pure CO2 production in situ allowing for CO2 sequestration. In a general chemical looping combustion process, a metal (Me) or the reduced metal oxide (MexOy-1) associated with its oxidized form (MexOy) is circulated between two reactorssthe oxidizer and reducer. In the reducer, the metal oxide (MexOy) reacts with the fuel to produce CO2, H2O, and metal (Me) or the reduced metal oxide (MexOy-1), as illustrated in eqs 1-1 and 1-2, respectively. Pure CO2 is ready for subsequent sequestration in the exit gas stream from the reducer after H2O is condensed. In the oxidizer, the metal (Me) or the reduced metal oxide (MexOy-1) reacts with air to form metal oxide (MexOy), as illustrated in eqs 2-1 and 2-2. The advantage of CLC compared to the normal combustion process is that CO2 is not mixed and diluted with nitrogen without any energy needed for separation. Other benefits include a large elimination of NOx emission9-11 and high thermal efficiency.12

(2n + m)MexOy + yCnH2m f (2n + m)xMe + myH2O + nyCO2 (1-1) or

(2n + m)MexOy + CnH2m f (2n + m)MexOy-1 + mH2O + nCO2 (1-2) or

xMe + y/2O2 f MexOy

(2-1)

MexOy-1 + 1/2O2 f MexOy

(2-2)

The metal oxides with their reduced metal oxides or metals, which are used as oxygen carriers in the CLC, must have sufficient reactivities in reduction and oxidation and enough strength to limit particle breakage and attrition. A number of (9) Richter, H.; Knoche, K. Reversibility of Combustion Processes. ACS Symp. Ser. 1983, 235, 71-86. (10) Podolski, W. F.; Swift, W. M.; Miller, S. A. Air Emissions from Pressurized Fluidized Bed Combustor, Pressurized Fluidized Bed Combustion; Cuenca, M. A., Anthony, E. J., Eds.; Chapman & Hall: London, 1995. (11) Jin, H.; Okamoto, T.; Ishida, M. Development of a Novel ChemicalLooping Combustion: Synthesis of a Looping Material with a Double Metal Oxide of CoO-NiO. Energy Fuels, 1998, 12 (6), 1272. (12) Ryu, H. J.; Bae, D. H.; Han, K. H.; Lee, S. Y.; Jin, G. T. Reactivity Study on Ni-based Oxygen Carrier Particle in a Fixed Bed Reactor for Chemical-Looping Combustor. Theor. Appl. Chem. Eng., KIChE 2002, 8, 1329. (13) Hatanaka, T.; Matsuda, S.; Hatano, H. A New-concept Gas-Solid Combustion System ‘MERIT’ for High Combustion Efficiency and Low Emissions. Proc. Intersoc. Energy ConVers. Eng. Conf. 1997, 30, 944948. (14) Herzog, H.; Eliasson, B.; Kaarstad, O. Capturing Greenhouse Gases. Sci. Am. 2000, 282 (2), 72-79. (15) Ishida, M.; Jin, H. A Novel Combustor Based on Chemical-looping Reactions and its Reaction Kinetics. J. Chem. Eng. Jpn. 1994, 27, 29630. (16) Ishida, M.; Jin, H. Novel Chemical-looping Combustor without NOx Formation. Ind. Eng. Chem. Res. 1996, 35, 2469-2472. (17) Ishida, M.; Jin, H.; Okamoto, T. A Fundamental Study of a New Kind of Medium Material for Chemical-looping Combustion. Energy Fuels 1996, 10, 958-963. (18) Ishida, M.; Jin, H.; Okamoto, T. Kinetic Behaviour of Solid Particle in Chemical-looping Combustion: Suppressing Carbon Deposition in Reduction. Energy Fuels 1998, 12, 223-229. (19) Ishida, M.; Yamamoto, M.; Saito, Y. Experimental Works on Innovative Chemical-looping Combustor. ECOS ‘99, Proceedings, International Conference on Efficiency, Costs, Optimization, Simulation and Environmental Aspects of Energy Systems, Tokyo, June 8-10, 1999; pp 306-310.

metals have been discussed in the literature,13-34 such as Fe, Ni, Co, Cu, Mn, and Cd, as well as some metal blends. The investigation of the oxidation and reduction kinetics of selected metal oxides has been intensively carried out in either a thermogravimetric analysis or lab-scale fixed bed and fluidized bed using gaseous fuels such as H2, CO, or CH4.13-34 The kinetics of reactions vary widely depending upon the type of metal oxide, particle size (70 µm to 2 mm), reduction gas (H2, CO, and CH4), and temperature (600-1000 °C). Generally, Cu, Ni, and Co and their oxides showed higher oxidation and reduction reactivities and greater durability after repeated oxidation and reduction cycles than those of Fe.21-24,28-31 It was also found that an impregnated type of oxygen carrier could increase the reactivity and durability, even with its particle size being as large as about 2 mm in diameter.21,23-27,30-32 An almost full conversion of the reactants could be achieved in minutes for the impregnated type of oxygen carriers. The candidates for the inert support materials could be SiO2, Al2O3, yttria-stabilized zirconium, TiO2, and MgO. However, different combinations of active materials of oxygen carriers with inert support materials showed different crushing strengths and sintering temperatures.28,29 During the experiments with CH4, carbon deposition may occur, which could cause a dramatic loss of reactivity. However, a high concentration of water vapor available in the reducer can help to eliminate carbon deposition.23,28,31 Starting from the introduction of CLC in 1983, the majority of studies and process development has been with gaseous fuels (20) Ishida, M.; Yamamoto, M.; Ohba, T. Experimental Results of Chemical-looping Combustion with NiO/NiAl2O4 Particle Circulation at 1200C. Energy ConVers. Manage. 2002, 43, 1469. (21) Jin, H.; Ishida, M. Reactivity Study on Natural-gas-fueled Chemicallooping Combustion by a Fixed Bed. Ind. Eng. Chem. Res. 2002, 41, 40044007. (22) Jin, H.; Okamoto, T.; Ishida, M. Development of a Novel Chemicallooping Combustion: Synthesis of a Looping Material with a Double Metal Oxide of CoO-NiO. Energy Fuels 1998, 12, 272-1277. (23) Jin, H.; Okamoto, T.; Ishida, M. Development of a Novel Chemicallooping Combustion: Synthesis of a Solid Looping Material of NiONiAl2O4. Ind. Eng. Chem. Res. 1999, 38, 126-132. (24) Cho, P.; Mattisson, T.; Lyngfelt, A. Comparison of iron-, nickel-, copper- and manganese-based oxygen carriers for chemical looping combustion. Fuel 2004, 83, 1215-1225. (25) Mattisson, T.; Johansson, M.; Lyngfelt, A. Multi-cycle reduction and oxidation of different types of iron oxide particles. Application to chemical-looping combustion. Energy Fuels 2004, 18, 628-637. (26) Johansson, M.; Mattisson, T.; Lyngfelt, A. Investigation of Fe2O3 with MgAl2O4 for chemical-looping combustion. Ind. Eng. Chem. Res. 2004, 43, 6978-6987. (27) Mattisson, T.; Ja¨rdna¨s, A.; Lyngfelt, A. Reactivity of some metal oxides supported on alumina with alternating methane and oxygens. Application for chemical-looping combustion. Energy Fuels 2003, 17, 643651. (28) Adanez, J.; de Diego, L. F.; Garcia-Labiano, F.; Gayan, P.; Abad, A. Selection of Oxygen Carriers for Chemical-Looping Combustion. Energy Fuels 2004, 18, 371-377. (29) de Diego, L. F.; Garcia-Labiano, F.; Adanez, J.; Gayan, P.; Abad, A.; Corbella, B. M.; Palacios, J. M. Development of Cu-based Oxygen Carriers for Chemical-Looping Combustion. Fuel 2004, 83, 1749-1757. (30) Mattisson, T.; Johansson, M.; Lyngfelt, A. The use of NiO as an oxygen carrier in chemical-looping combustion. Fuel 2005, submitted for publication. (31) Cho, P.; Mattisson, T.; Lyngfelt, A. Carbon Formation on Nickel and Iron Oxide-Containing Oxygen Carriers for Chemical-Looping Combustion. Ind. Eng. Chem. Res. 2005, 44, 668-676. (32) Mattisson, T.; Lyngfelt, A.; Cho, P. The use of iron oxide as an oxygen carrier for chemical-looping combustion of methane with inherent separation of CO2. Fuel 2001, 80, 1953-1962. (33) Wolf, J.; Anheden, M.; Yan, J. Performance Analysis of Combined Cycles with Chemical Looping Combustion for CO2 Capture. Proceedings of the 18th Pittsburgh Coal Conference, New South Wales, Australia, Dec. 4-7, 2001 [CD-ROM, session 23]. (34) Mattisson, T.; Jarders, A.; Lyngfelt, A. Reactivity of some Metal Oxide Support on Alumina with Alternating Methane and Oxygenapplication for Chemical-looping Combustion. Energy Fuels 2003, 17, 643651.

1838 Energy & Fuels, Vol. 20, No. 5, 2006

such as methane and natural gas.35-37 However, the methane and the natural gas supply cannot fully support the energy needs of the country’s electricity demand for the long term. Solid fuels such as coal and biomass have been seldomly used in the concept of chemical looping combustion because of their technical problems.41-43 Although the adaptation of CLC to the combustion of solid fuels presents many challenges, this adaptation is very attractive because of rich coal deposits. The control of other pollutants emitted in coal combustion, such as sulfur oxides and trace metals, can be conducted in an efficient way because of the very low volume of nitrogen-free flue gas that is generated in the reducer. The meaning is significant for sustaining the long-term utilization of rich coal resources in an economical, efficient, and environmental way. This paper is the first in a series of two, where we present the concept of a CLC process of solid fuels using a circulating fluidized bed (CFB) with three loop seals and the establishment of a theoretical frame of oxygen transfer capability, reaction enthalpy, chemical equilibrium, and kinetics for oxygen carriers of interest. In the second part of the series, we will evaluate the reduction reaction of selected Cu-based oxygen carriers by coal and other “opportunity solid fuels” using simultaneous differential scanning calorimetry-thermogravimetric analysis to simulate a microreactor, an X-ray diffractometer and a scanning electron microscope as the characterization methods for the solid reaction residues, and thermogravimetric analysis coupled with mass spectrometry to characterize the evolved gas compositions. 2. Technical Approach for CLC of Solid Fuels 2.1. Process Analysis and Technical Issues. There are two approaches to applying CLC to solid fuel combustion. The first approach is to gasify solid fuels in a separate gasifier with pure oxygen to produce a syngas of CH4, CO, and H2 without nitrogen. Syngas is supplied for the CLC system, which is similar to the CLC process using a natural gas with an additional gasifier. However, the production of pure oxygen and the fabrication of an additional gasifier will be required and, thus, dramatically increase the capital cost of the CLC system. Approach 2, which is the proposed CLC process for solid fuels in this study, is to directly supply solid fuel into the reducer (35) Lyngfelt, A.; Leckner, B.; Mattisson, T. A fluidized-bed combustion process with inherent CO2 separation; application of chemical-looping combustion. Chem. Eng. Sci. 2001, 56, 3101-3113. (36) Zafar, Q.; Mattisson, T.; Gevert, B. Integrated Hydrogen and Power Production with CO2 Capture Using Chemical-Looping Reformings. Redox Reactivity of Particles of CuO, Mn2O3, NiO, and Fe2O3 Using SiO2 as a Support. Ind. Eng. Chem. Res. 2005, 44 (10), 3485-3496. (37) Mattisson, T.; Lyngfelt, A. Applications of chemical-looping combustion with capture of CO2. Second Nordic Minisymposium on CO2 Capture and Storage, Go¨teborg, Sweden, 2001. (38) U.S. DOE Project Fact Sheet, Hybrid Combustion-Gasifictaion Chemcial Looping Coal Power Technology Development. www.netl.doe.gov/coal (accessed Dec 2005). (39) Andrus, H. E., Jr.; Chiu, J. H.; Stromberg, P. T.; Thibeault, P. R. Alstom Power Inc, Alstom’s Hybrid Combustion-Gasifictaion Chemcial Looping Coal Power Technology Development. Twenty-Second Annual International Pittsburgh Coal Conference, Pittsburgh, PA, Sept 12-15, 2005. (40) Bedick, R. C. Advancements in Coal-fired Power Generation Technologies. Coal Utilization Technologies Workshop, National Research Center for Coal & Energy, Morgantown, WV, Sept 22, 2004. (41) Cao, Y.; Cheng, Z.; Meng, L.; Riley, J. T.; Pan, W.-P. Reduction of Solid Oxygen Carrier (CuO) by Solid Fuel (Coal) in Chemical Looping Combustion. Prepr. Pap.sAm. Chem. Soc., DiV. Fuel Chem. 2005, 51, 99100 (42) Cao, Y.; Pan, W.-P. Chemical Reversal Cycle of Solid Oxygen Carrier for Producing Pure Oxygen or Oxygen Rich Gas Stream. Prepr. Pap.sAm. Chem. Soc., DiV. Fuel Chem. 2005, 51, 415-416 (43) Cao, Y.; Riley, J. T.; Pan, W.-P. Application of a circulating fluidized bed process for the chemical looping combustion of solid fuels. Prepr. Pap.sAm. Chem. Soc., DiV. Fuel Chem. 2004, 49 (2), 815-816.

Cao and Pan

in the CLC system. It has economic advantages, but several technical problems. There are two reaction paths in this approach, direct reduction of the oxygen carrier by solid fuels and indirect reduction by syngas from solid fuel gasification in the reducer. There is actually a mixed mechanism of these two reaction paths mentioned above. The actual reaction modes occurring in the chemical looping combustion of solid fuels are dependent on the physical properties, reaction thermodynamics, and kinetics of the selected solid fuels and oxygen carriers. For the first path occurring in the reducer, the primary technical concern is the lower reactivity between solid fuel and the oxygen carrier due to low solid-solid contact efficiency. For the second path, the technical concern is the lower solid fuel gasification rate compared with its combustion rate at the same evaluated temperature. Such issues need to prolong the residence time of solid fuels inside the reducer to fulfill the higher carbon conversion efficiency. Therefore, reduction mechanisms of oxygen carriers by solid fuels and their kinetics in the reducer of the chemical looping combustion process should be examined more closely. Other technical issues pertaining to the development of the chemical looping combustion of solid fuels also include (1) the separation of the oxygen carrier from unburned carbon and fly ash, (2) the prevention of unburned carbon particles in the oxidizer, (3) the energy distribution between the oxidizer and reducer in the presence of an endothermic gasification process, (4) the recycling of solid phase materials and system pressure balance, and (5) the prevention of gas leakage between the reducer and oxidizer. All technical issues will be addressed and discussed in this study, except issue 5, which has been intensively discussed in ref 44. Presently, the only available technology to use the concept of chemical loops to combust the solid fuel is the Alstom Power Inc.’s Hybrid Combustion-Gasification Chemical Looping Coal Power Technology.38-40 It applies two chemical loops of CaSO4-CaS and CaCO3-CaO and also one thermal loop with bauxite. In the oxidizer of the first chemical loop of CaSO4CaS, coal was reacted with CaSO4 to produce CO, not CO2 directly. CO needs to be further converted into CO2 through a shift reaction and then concentrated for sequestration through the second chemical loop of CaCO3-CaO. In this process, one more chemical loop, one more thermal loop, and one shift reaction step are introduced into the whole system to conduct solid fuel combustion and CO2 concentration simultaneously. Thus, its economic competitiveness and process complexity need to be further improved. 2.2. The Concept of the Proposed Process. To overcome the technical difficulties mentioned above in an economical way, we propose a CFB with a combination of several loop seals for the chemical looping combustion of solid fuels, as briefly illustrated in Figure 1. In this approach, the facility will consist of three major componentssa high-velocity riser acting as an oxidizer of the oxygen carrier, a down-flow moving bed or a bubbling fluidized bed acting as a reducer of the oxygen carrier in which solid fuels are fed, and a low-velocity bubbling bed or a turbulent fluidized bed acting as a deep reduction reactor and separator for the oxygen carrier from fly ash as well as unburned carbon. Among them, the reducer and separator are combined together to assemble a larger loop seal. The other two small loop seals will act as connectors of solid recycling between the oxidizer and the reducer with the separator. The gas velocity in the riser provides the driving force for the (44) Johansson, E.; Lyngfelt, A.; Mattisson, T.; Johnsson, F. Gas Leakage Measurements in an Interconnected Fluidized Bed for Chemical-looping Combustion. Powder Technol. 2003, 134, 210-217.

Chemical Looping Combustion by Solid Fuels

Energy & Fuels, Vol. 20, No. 5, 2006 1839

Figure 1. System configuration for the pilot-scale chemical looping combustion process of solid fuels.

circulation of particles between the three components. The loop seals balance the system pressure and allow solids to flow from locations of lower static pressure to those of higher static pressure. A heat exchanger can be installed in the air oxidizing riser to produce steam for power generation. In the riser, air is supplied as the fluidizing agent to oxidize the metal or the reduced metal oxide to its oxidized form where oxygen is transferred from the air to the oxygen carrier. After being separated by a cyclone, the oxidized oxygen carrier returns to the reducer. Steam or recycling CO2 is used as the aeration or transport media and gasification agent in the reducer. Metal oxide is reduced to elemental metal or the reduced metal oxide in the reducer by direct or indirect reaction with solid fuels, which is dependent on the properties and reaction mechanisms of the solid fuel and oxygen carrier. In the bubbling bed or the turbulent fluidized bed, deep reduction of the oxygen carrier, deep conversion of the solid fuels, and full separation of the reduced oxygen carrier with fly ash will occur simultaneously. The reduced oxygen carrier will return to the riser (oxidizer) via another small loop seal. The remaining unburned carbon is recycled to the reducer or directly combusted in another facility because of its lower reactivity. Fresh oxygen carrier probably has to be added to maintain the reactivity of the oxygen carrier. The volume of the gas flow in the oxidizer is much larger than that in a reducer because a large amount of nitrogen is carried in by the air. The oxidizing rate of the oxygen carrier is

much faster than the rate of the reducing reaction. Thus, a high gas velocity has been chosen in the oxidizing reactor to keep a reasonable size of reactor in consideration of the capital cost. In the reducer, relatively long residence times, on the order of tens of minutes, are needed for the reduction of the oxygen carrier directly by solid fuels or indirectly by syngas from gasification of the solid fuels in the same reactor. Only a small amount of gas is needed for fossil fuel gasification, so a downflow moving bed or a bubbling fluidized bed is selected. The ash with unburned carbon produced is much lighter (800-1200 kg/m3 in particle density) compared to the oxygen carrier (generally above 5000 kg/m3 in particle density, which is dependent on its porosity). Thus, the oxygen carrier and fly ash can be separated on the basis of density difference. By controlling the operating velocity, fly ash will be entrained out of a bubbling fluidized bed or a turbulent fluidized bed with “flue gas 2” and enter “cyclone 2” for cleanup. The flue gas exiting the bubbling bed is a nitrogen-free gas with CO2, H2O, and possibly CO and H2, which depends on the properties of the selected oxygen carriers. The system pressure balance is essential to the buildup of solid recycling and the adjustability of the recycle rate. The actual pressures at specific locations of the reactor for several operating conditions form pressure profiles, which are shown in Figure 2. In this figure, points 4-8, points 8-12, and points 12-16 represent the upper loop seal, the middle loop seal

1840 Energy & Fuels, Vol. 20, No. 5, 2006

Cao and Pan

Figure 2. Pressure profiles of the chemical looping combustion system of solid fuels.

(reducer and separator), and the down loop seal, respectively. The pressure difference is distributed in these loop seals to fulfill the task of transporting recycled solids from a lower pressure at point 4 to a higher pressure at point 8, then from point 8 (lower pressure) to point 12 (higher pressure), and further from point 12 (lower pressure) to point 17 (higher pressure). Generally speaking, the higher recycle rate of solids will make the pressure differential inside the riser increase; therefore, a higher pressure difference created by the loop seals must be made in the CLC system. To adjust the pressure difference, one needs to adjust the gas velocity inside the loop seals so that the flow pattern inside the loop seals will be changed, especially in the middle loop seal. The capability of loop seals for adjusting pressure differences to adapt to changes in operation conditions is dependent on the operation modes of the loop seals and their sizes. For the same solid recycle rate, the larger loop seals or that made with the fluidized bed mode is always better than the smaller loop seals or that made with the moving bed. In the one design mode, the middle loop seal is made up of a moving bed on one side, which acts as the reducer, and is combined with a bubbling fluidized bed, which acts as a separator for ash and the oxygen carrier. The pressure difference in the moving bed is always smaller, as shown in points 8′ and 9′ in Figure 2 (red for the high mass flow rate and green for the low mass flow rate operating conditions). If solids flow is intended to be operated at the higher mass flow rate condition, the upper and the down loop seals must work at the larger pressure difference. However, these two loop seals are both

small and cannot create the large pressure difference. The static pressure at the point 5′ location, which is expected, will actually be at the point 5′′ location in the dashed line, which may actually be lower than that at the point 8′ location. The normal pressure profiles in the upper loop seal may collapse. Then, the solids flow may back-flow into the upper loop seals, which would cause the cycle system of the CLC to be unsustainable. If recycling of the solid is operated at a lower mass flow rate condition, as shown in the pressure profile marked with green, it is very difficult to control and adjust the system under such narrow pressure differences between loop seals. In the preferable design mode, the pressure profile of the middle loop seal was indicated by the black line in Figure 2. The reducer works under fluidized flow conditions, and the separator works under turbulent flow conditions. In this configuration, the larger loop seal can work at larger pressure differences (points 8-9 in black) and, thus, let small loop seals (upper and down) work reasonably at a smaller pressure difference. This modified system also will provide the flexibility of control of the mass flow rate (higher or lower). Thus, it is suitable for system setup, startup, stabilization, and condition testing to occur at the pilot-scale stage. An additional benefit is that bubble fluidization could provide good solid-to-solid contact between solid fuels and the oxygen carrier in the reducer, rather than in the moving bed. The operating gas velocity at the different reactors could be 5 times the terminal velocity of the selected oxygen carrier in the riser to maintain the fast fluidization state, 5 times the minimum fluidization velocity of

Chemical Looping Combustion by Solid Fuels

Energy & Fuels, Vol. 20, No. 5, 2006 1841

Table 1. Physical Properties and Oxygen Transfer Capability of Oxygen Carriers45

No.

reduction reaction

melting point of the reduced metal form, °C

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

2CuO + C ) 2Cu + CO2 2Cu2O + C ) 4Cu + CO2 2NiO + C ) 2Ni + CO2 2Co3O4 + C ) 6CoO + CO2 1/ Co O + C ) 3/ Co + CO 2 3 4 2 2 2CoO + C ) 2Co + CO2 6Mn2O3 + C ) 4Mn3O4+ CO2 2Mn2O3 + C ) 4MnO + CO2 2/ Mn O + C ) 4/ Mn + CO 3 2 3 3 2 2Mn3O4 + C ) 6MnO + CO2 1/ Mn O + C ) 3/ Mn + CO 2 3 4 2 2 2MnO + C ) 2Mn + CO2 6Fe2O3 + C ) 4Fe3O4 + CO2 2Fe2O3 + C ) 4FeO + CO2 2/ Fe O + C ) 4/ Fe + CO 3 2 3 3 2 2Fe3O4 + C ) 6FeO + CO2 1/ Fe O + C ) 3/ Fe + CO 2 3 4 2 2 2FeO + C ) 2Fe + CO2 2PbO + C ) 2Pb + CO2 2CdO + C )2Cd + CO2

1083 1083 1452 1480 1480 1480 1564 1650 1260 1650 1260 1260 1538 1420 1275 1420 1275 1275 327.5 320.9

melting point of the oxidized metal form, °C

specific density of the reduced metal form FR, kg/m3

specific density of the oxidized metal form FO, kg/m3

moles of metal per mole of oxygen transfer (N), mol/mol

(NxFR)M/ (NxFR)Cu in reducer

(NxFO0.5)M/ (NxFO0.5)CuO in oxidizer

1026 1235 1452 895 895 1800 1080 1080 1080 1564 1564 1650 1560 1560 1560 1538 1538 1420 886 900

8920 8920 8900 8900 8900 8900 4856 5180 7200 5180 7200 7200 5200 5700 7030 5700 7030 7030 11340 8650

6450 6000 7450 6070 6070 5680 4810 4810 4810 4856 4856 5180 5120 5120 5120 5200 5200 5700 8000 8150

1 2 1 3 0.75 1 6 2 0.67 3 0.75 1 6 2 0.67 3 0.75 1 1 1

1.0 2.0 1.0 3.0 0.7 1.0 3.3 1.2 0.5 1.7 0.6 0.8 3.5 1.3 0.5 1.9 0.6 0.8 1.3 1.0

1.0 1.9 1.1 2.9 0.7 0.9 5.2 1.7 0.6 2.6 0.7 0.9 5.3 1.8 0.6 2.7 0.7 0.9 1.1 1.1

the selected oxygen carrier in the reducer to maintain the bubbling fluidization state, and the terminal velocity of the fly ash in the separator to maintain the turbulent fluidization state. 2.3. The Oxygen Carriers for the CLC of Solid Fuels and Thermodynamic Analysis. On the basis of the oxygen transfer capability, an energy balance analysis, and a thermodynamics analysis, copper (Cu) seems to be the better choice as an oxygen carrier for the CLC system of solid fuels. 2.3.1. Physical Properties of Oxygen Carriers. As described in the section on process analysis in this paper and in previous studies,35,41-43 the interconnection of the fast fluidized bed with the another fluidized bed or the moving bed is the most possible reactor configuration to conduct the CLC process. For a reactor in which solid particles are flowing, moving, and recirculating, the possibility of particle agglomeration should be avoided. For metal-based oxygen carriers, the melting points of the selected metals, reduced metal oxides, and metal oxides are important parameters in evaluating their agglomeration tendencies. The melting of oxygen carriers may also resul in a loss of reactivity. The melting points of candidate metals, their reduced metal oxides, and metal oxides are shown in Table 1.45 The majority of metals and their metal oxides have very high melting points, which are always higher than 1200 °C, except those with melting points around 1050 °C, which are within the operation temperature range of the CLC process, such as Cu, CuO, and Mn2O3, which are in italics. Some metals and their metal oxides have very low melting points such as PbO, CdO, and Co3O4, which are in bold and must be removed from the candidate list to be oxygen carriers in the CLC process. However, the operating temperature of the CLC process is also dependent on the acceptable reactivity of the oxygen carriers. Previous studies indicated that Cu-based oxygen carriers had a reactivity of 100% reduction within minutes at low temperatures (600-900 °C)19,34-36 in the fixed bed or fluidized bed testing facilities. When copper oxide was doped on the substrate, it was indicated that most of its oxygen was active for reaction with methane, and the highest efficiency could be achieved with reduction rates up to 100%/ min and oxidation rates up to 25%/min.24,28-29 2.3.2. Oxygen Transfer Capability of the Oxygen Carrier. The utilization efficiency of an oxygen carrier is dependent on its oxygen transfer capability, which is listed in Table 1 as the (45) Perry, R. H.; Green, D. W. Perry’s Chemical Engineers’s Handbook, 7th ed.; The McGraw-Hill Companies, Inc.: New York, 1997

moles of metal per mole of oxygen transfer (N). The recirculation of the oxygen carrier required energy consumption, which is proportional to the pressure drop across the reactor.35 In the oxidizer, which is a fast fluidization bed, the pressure drop is proportional to the critical velocity labelled as the terminal velocity Ut.46 In the reducer, which is a bubble fluidization bed or a moving bed, the pressure drop is proportional to the critical velocity labelled as the minimum fluidizing velocity Umf.46 As shown in eqs 3 and 4, Fp, which is the particle density of the oxygen carriers, is the most important parameter affecting Ut and Umf. All other parameters in eqs 3 and 4 can be controlled during the manufacturing of the oxygen carriers and system operation parameters. In eq 3, the factor value of Fp is 0.5, and in eq 4, the factor value of Fp is 1. These two parameters can be multiplied by N relative to the same parameters of the CuO-Cu system, which are (NxFp)/(NxFp)Cu-CuO or (NxFp)/ (NxFp0.5)Cu-CuO, to evaluate the relative energy consumption by using different oxygen carriers when the oxygen carriers are directly manufactured by their metal or metal oxide. Some oxygen carriers were manufactured as an impregnated type. In this case, N can be used as the relative energy consumption by using different oxygen carriers.

Ut ) [4dp(Fp - Fg)g/(3FgCD)]1/2

(3)

As shown in Table 1, N values and (NxFp)/(NxFp)Cu-CuO or

Umf ) [dp2(Fp - Fg)g(mf3Φs2)]/[(150µ)(1 - mf)]

(4)

(NxFp)/(NxFp0.5)Cu-CuO values follow the same trend, so we can focus on the N value to evaluate energy consumption. The best candidates for oxygen carriers should be those with a smaller N value. The minimum value of N is 0.67 for reactions 9 and 15 and the next is 0.75 for reactions 5, 11, and 17 for Co-, some Mn-, and some Fe-based oxygen carriers. N values of other reactions are close to 1 for Cu-, Ni-, and some Co- and Febased oxygen carriers. In Table 1, we listed reactions in bold if the N value was greater than 3 and in italics for N values of 1-3. By evaluating the parameters of (NxFp)/(NxFp)Cu-CuO and (NxFp)/(NxFp0.5)Cu-CuO, the same conclusion can be made. (46) Kunii, D.; Levenspiel, O. Fluidization Engineering, 2nd ed.; Butterworth-Heinemann: Woburn, MA, 1990; Butterworth-Heinemann Series in Chemcial Engineering, ISBN: 0-409-90233-0.

1842 Energy & Fuels, Vol. 20, No. 5, 2006

Cao and Pan

Table 2. Enthalpies of Reduction Reaction by Carbon at 1000 °C and 1 atm (Calculations Based on Data from ref 45) C + O2 f CO2, -392.75 kJ/mol endothermic

exothermic

2NiO + C f 2Ni + CO2, 75.21kJ/mol 2CoO + C f 2Co + CO2, 73.92 kJ/mol 1/ Co O + C f 3/ Co + CO , 53.9 kJ/mol 2 3 4 2 2 2/ Mn O + C ) 4/ Mn + CO , 239.61 kJ/mol 3 2 3 3 2 6Mn3O4 + C ) 4MnO + CO2, 54.21 kJ/mol 1/ Mn O + C f 3/ Mn + CO , 296.65 kJ/mol 2 3 4 2 2 2MnO + C f 2Mn + CO2, 378.98 kJ/mol 6Fe2O3 + C f 4Fe3O4 + CO2, 83.56 kJ/mol 2Fe2O3 + C f 4FeO + CO2, 158.40 kJ/mol 2/ Fe O + C f 4/ Fe + CO , 146.37 kJ/mol 3 2 3 3 2 2Fe3O4 + C ) 6FeO + CO2, 195.78 kJ/mol 1/ Fe O + C ) 3/ Fe + CO , 151.27 kJ/mol 2 3 4 2 2 2FeO + C ) 2Fe + CO2, 136.44 kJ/mol

2CuO + C f 2Cu + CO2, -96.51 kJ/mol 2Cu2O + C f 4Cu + CO2, -61.04 kJ/mol 6Co3O4 + C ) 6CoO + CO2, -8.63 kJ/mol 6Mn2O3 + C ) 4Mn3O4 + CO2, -216.63 kJ/mol 2Mn2O3 + C ) 4MnO + CO2, -36.07 kJ/mol

Any candidate in italics or bold is unsuitable for use as an oxygen carrier in a CLC system; therefore, any candidates not in italics or bold have passed the selection criteria. They include Ni-based (NiO-Ni), Co-based (CoO-Co), Mn-based (Mn3O4Mn and MnO-Mn), and Fe-based (Fe2O3-Fe, Fe3O4-Fe, and FeO-Fe) oxygen carriers. Because of the perfect reactivity of Cu-based oxygen carriers at lower temperatures, reaction 1 is still included in the candidates list for the CLC process of solid fuels. It was reported that Fe-based oxygen carriers have a lower reactivity than those of other selected oxygen carriers. Aluminadoped manganese-based oxides could produce stable nonreactive compounds of spinel (MnAl2O4 and Mn2AlO4), which may result in a considerable loss of the oxygen transfer capability of oxygen carriers.34 Ni-based (NiO-Ni), Co-based (CoO-Co), and Cu-based (CuO-Cu) carriers are expected to be good candidates for oxygen carriers. Energy Balance Analysis. The oxygen carrier reduction process by solid fuel is far more complicated compared to that by gaseous fuels. The process is governed by the prevailing chemical thermodynamics and kinetics. If the indirect path (the gasification of solid fuels followed by the reduction of metal oxides by gaseous gasification products) is dominant, the properties of pyrolysis and gasification of the solid fuels should be considered. All reactions related to the pyrolysis and gasification of fossil fuels are endothermic, as illustrated by eqs 5-7. All enthalpy data are based on 1 atm and 25 °C.45 In conventional gasification technologies, the self-combustion of solid fuels is needed to provide the heat requirement of the endothermic pyrolysis and gasification processes. This is called the autothermal process. In the proposed CLC system, the gasification process for solid fuel occurs simultaneously with a reduction process for the oxygen carrier in the reducer where no oxygen exists. There are two ways to supply heat for the solid fuel gasification process. One is the reduction of the oxygen carrier indirectly by solid fuels or their product gases, and the other method is to use a heat-transfer material with a high heat capacity to transfer heat from the oxidizer to the reducer. Obviously, the previous one is the better choice to prevent additional energy consumption by recycling of the heattransfer material.

CnHmOp f aCO2 + bH2O + cCH4+ dCO + eH2 + f(C2C5), 20.9 kJ/mol (5) Until now, just a few metal oxides that have been examined

C + H2O f CO + H2, 118.3 kJ/mol

(6)

C + CO2 f 2CO, 160.5 kJ/mol

(7)

as oxygen carriers for the CLC process have shown exothermic

properties when reacted with carbon or syngas. The possible reactions related to oxygen carrier candidates and carbon in the reducer are shown in Table 2. There are two steps (oxidizing and reducing) in the process of chemical looping for solid fuel combustion. Thus, the enthalpy of solid fuel combustion in the chemical looping process is equal to the sum of the enthalpies of the two steps. For the different oxygen carriers shown, there is a different distribution of enthalpies in two steps, and even the endothermic and exothermic reactions may be changed. Cubased and CoO-Co oxygen carriers are the only choices whose reduction is exothermic due to the smaller enthalpy of the oxidation compared to the direct combustion of solid fuel. The Mn3O4-MnO oxygen carrier should be removed from the suggested candidate list because of the reason (N ) 6) previously mentioned. Despite endothermic properties in the reducer, oxygen carriers such as NiO-Ni, CoO-Co, Co3O4-Co, Mn3O4-MnO, and Fe2O3-Fe3O4 have smaller reaction enthalpies, so they can be included as candidates for oxygen carriers if heat-transfer materials will be used for heat transfer in the reducer. Because of the reason previously mentioned (melting points and N values), Co3O4-Co, Mn3O4-MnO, and Fe2O3Fe3O4 can be removed from the suggested candidate list. Thermodynamics Analysis. Chemical reaction thermodynamics are important for the control of CO2 purity. From the standard Gibbs free energy changes, the equilibrium constants can be calculated for the various reactions of metal oxide reduction and solid fuel gasification for a wide range of operating temperatures. The phase diagrams for the reduction reactions using the reducing agents CO and H2 produced from the solid fuel gasification process at atmospheric pressure are shown in Figures3 and 4, respectively. Figure 3 shows an equilibrium gas ratio of PCO2/PCO as a function of the temperature for the reduction of various metal oxides. It shows that the ratio varies from approximately 105 for the reduction of CuO to Cu, Cu2O to Cu, Mn3O4 to MnO, and Fe2O3 to Fe3O4 to values on the order of 10-5 or less for the reduction of MnO to its elemental state of Mn. Information related to the Boudouard reaction, as illustrated in eq 5, is also presented in Figure 3. For the reduction of a metal oxide with CO in the absence of solid carbon, the oxides of copper and nickel will be reduced to their elemental forms at gas ratios between 105 and 102. Therefore, the completeness of the reaction could be achieved and a highly concentrated CO2 stream will be obtained in the proposed CLC system. On the other hand, the reduction of MnO requires a CO2-free environment, which is practically impossible for the CLC of solid fuel to achieve. Because carbon is also present in the reaction mixture in the proposed reducer, the reduction of metal oxide and carbon gasification to CO occurs simultaneously. As shown in Figure 3, above the simultaneous equilibrium temperature, where two

Chemical Looping Combustion by Solid Fuels

Figure 3. Variation of the thermodynamic equilibrium factor for MexOy-CO as a function of temperature (calculations based on data from ref 45).

Figure 4. Variation of the thermodynamic equilibrium factor for MexOy-H2 as a function of temperature (calculations based on data from ref 45).

curves for the reduction of metal oxides and the Boudouard reaction intersect, Fe3O4 will be converted to FeO above 650 °C and then to Fe above 700 °C at a low PCO2/PCO ratio of ∼10. In contrast, the curves for CuO, Cu2O, NiO, and Fe2O3 do not intersect with the carbon curve, even at the temperature range of interest (600-1200 °C). The simultaneous reactions are not limited by thermodynamics; they are determined entirely by kinetics. In this case, the gas constituents produced will have an intermediate impact on the value of the PCO2/PCO ratio. Generally, the reduction reaction of the metal oxide is faster than the solid fuel gasification in the CLC process. H2 is another product from the solid fuel gasification process using H2O as the gasification agent. Figure 4 is the thermodynamics diagram for the PH2O/PH2 equilibrium system, similar to the PCO2/PCO system presented in Figure 3. The reduction of metal oxides with H2 is less exothermic than its corresponding reaction with CO. Moreover, equilibrium of the water-gas reaction will occur and shift to the right at lower temperatures. This shows that H2 at high temperatures is a better reducing agent than CO for oxygen carriers. Solid fuel pyrolysis and gasification may produce some CH4 in the reducing stream. Thermodynamics shows that CH4 could be a better reducing agent than either CO or H2 at high temperatures. The tendency

Energy & Fuels, Vol. 20, No. 5, 2006 1843

of CH4 decomposition to result in oxygen carrier deactivation by carbon deposition may be largely eliminated under higher partial pressures of H2O and CO2 in the reducer. The main constituent of solid fuels is carbon. The possibility of directly reducing metal oxides with carbon can also be calculated using thermodynamic theory.45 All of these reduction reactions with carbon are heterogeneous reactions with only one gaseous species included. On the basis of thermodynamics theory, if the gas (CO2) is ideal, the equilibrium constant expressions for these reactions cannot include the solid phase because of the fact that pure solid phases are nearly equal to unity for moderate pressures. Thus, a phase diagram relating the partial pressure of CO2 and the temperature can be constructed as shown in Figure 5. In this figure, the equilibrium line for each oxygen carrier divides the 2-D area of temperature and CO2 partial pressure (PCO2) into two zones in the temperature range of interest. Metal oxides, reduced metal oxides or metals, carbon, and CO2 are present simultaneously only at the equilibrium line correlating the reaction temperature and CO2 partial pressure. Above this equilibrium line, metal oxides and carbon can have a stable existence. Under this equilibrium line, metal or reduced metal oxides can have a stable existence. The temperature and CO2 partial pressure are the parameters that control the extent of the reaction and its direction. Figure 5 indicates that the CO2 partial pressure (PCO2) of Cu-based, Nibased, and Co-based oxygen carriers can reach above 103 so that 99.999% purity of the CO2 stream can be assured in view of the reduction of a metal oxide directly by carbon in the temperature range of interest. Above 1000 °C, Fe-based oxygen carriers can only ensure the PCO2 to be above 10 with a CO2 stream purity of 90%. Generally, Mn-based oxygen carriers lack practicality because of a very low PCO2, at about 10-7 for the MnO-Mn reaction system and 10-3 for the Mn3O4-MnO reaction system. It is interesting to find that Mn2O3-Mn3O4 has a higher equilibrium PCO2 at about 1017, but it still needs to be rejected as an oxygen carrier because of a high N value, which was mentioned previously. 3. Conclusion In this study, a new CLC process of solid fuels, which is based on one chemical loop by metal and its oxide, is proposed, and its technical issues are intensively discussed, including the selection of an optimum reactor and oxygen carriers. On the basis of the requirement of recycling the oxygen carrier between the oxidizer and the reducer and the properties of the reaction occurring inside the reactors, an interconnected CFB with three loop seals was envisioned. The riser of the CFB was used as the oxidizer, and a larger loop seal was used as the reducer and separator, which was operated in bubbling fluidization or moving bed mode for the reducer and in turbulent fluidization mode for the separation of ash and the oxygen carrier. The other two small loop seals were used for system pressure balance. On the basis of the physical properties of oxygen carriers and a thermodynamics analysis of the reduction reaction between oxygen carriers and solid fuels, CuO-Cu-, NiO-Ni-, and CoO-Co-based oxygen carriers were determined to be the optimal oxygen carriers for development of the CLC of solid fuels. They all have larger oxygen transfer capabilities. Cubased oxygen carriers are the only choice that may make the reducer self-sustaining or autothermal with its exothermic properties during reduction in these three candidates of oxygen carriers. The exothermic enthalpy property of the Cu-CuO oxygen carrier in its reduction process would be beneficial for

1844 Energy & Fuels, Vol. 20, No. 5, 2006

Cao and Pan

Figure 5. Variation of the thermodynamic equilibrium factor for MexOy-C as a function of temperature (calculations based on data from ref 45).

simplifying the operation of the reducer. The agglomeration tendencies of Cu-based oxygen carriers can be eliminated by decreasing the operating temperature because of its high reactivity even at the temperature range of interest in the reducer (600-900 °C). Thermodynamic analysis indicated that CO2 can be concentrated and purified to at least 99% purity for the gassolid reaction mode (reduction of the oxygen carrier by gasification products such as CO and H2) and even higher for the solid-solid reaction mode (reduction of the oxygen carrier directly by solid fuels) on the basis of the selected oxygen carriers. Generally, Mn-based oxygen carriers have several disadvantagesslower oxygen transfer capability (Mn2O3-Mn3O4, Mn2O3MnO, and Mn3O4-MnO), thermodynamic limitations of purify

ing the CO2 stream (Mn3O4-MnO and MnO-Mn), a lower melting point (Mn3O4), or a larger endothermic reduction enthalpy (Mn2O3-Mn, Mn3O4-Mn, MnO-Mn, and Mn3O4MnO). Fe-based oxygen carriers have the disadvantage of a larger endothermic reduction enthalpy and lower reactivity. Acknowledgment. This paper was prepared by the Institute for Combustion Science and Environmental Technology of Western Kentucky University with support, in part, by grants made possible by the U.S. Department of Energy SBIR program (DE-FG0204ER84036). The authors also acknowledge Dr. Kunlei Liu of the University of Kentucky (CAER) for providing good suggestions and comments in this study. EF050228D