10200
Ind. Eng. Chem. Res. 2010, 49, 10200–10211
Chemical Looping Technology and Its Fossil Energy Conversion Applications Liang-Shih Fan* and Fanxing Li Department of Chemical and Biomolecular Engineering, The Ohio State UniVersity, Columbus, Ohio 43210
The concept of chemical looping reactions has been widely applied in chemical industries, for example, the production of hydrogen peroxide (H2O2) from hydrogen and oxygen using 9,10-anthraquinone as the looping intermediate. Fundamental research on chemical looping reactions has also been applied to energy systems, for example, the splitting of water (H2O) to produce oxygen and hydrogen using ZnO as the looping intermediate. Fossil fuel chemical looping applications had been used commercially with the steam-iron process for coal from the 1900s to the 1940s and had been demonstrated at a pilot scale with the carbon dioxide acceptor process in the 1960s and 1970s. There are presently no chemical looping processes using fossil fuels in commercial operation. A key factor that hampered the continued use of these earlier processes for fossil energy operation was the inadequacy of the reactivity and recyclability of the looping particles. This factor led to higher costs for product generation using the chemical looping processes, compared to the other processes that use particularly petroleum or natural gas as feedstock. With CO2 emission control now being considered as a requirement, interest in chemical looping technology has resurfaced. In particular, chemical looping processes are appealing because of their unique ability to generate a sequestration-ready CO2 stream while yielding high energy conversion efficiency. Renewed fundamental and applied research since the early 1980s has emphasized on improvements over earlier shortcomings. New techniques have been developed for direct possessing of coal or other solid carbonaceous feedstock in chemical looping reactors. Significant progress demonstrated by the operation of several small pilot scale units worldwide indicates that the chemical looping technology may become commercially viable in the future for processing carbonaceous fuels. This perspective article describes the fundamental and applied aspects of modern chemical looping technology that utilizes fossil fuel as feedstock. It discusses chemical looping reaction thermodynamics, looping particle selection, reactor design, and process configurations. It highlights both the chemical looping combustion and the chemical looping gasification processes that are at various stages of the development. Opportunities and challenges for chemical looping process commercialization are also illustrated. 1. Introduction Energy is the backbone of modern society. A clean, relatively cheap, and abundant energy supply is a prerequisite for the sustainable economic and environmental prosperity of society. With the significant economic growth in the Asia Pacific region and the expected development in Africa, the total world energy demands are projected to increase from 462.4 quadrillion BTU in 2005 to well over 690 quadrillion BTU by 2030.1 The projected energy supply through 2030 will be drawn from fossil fuels (i.e., oil, coal, and natural gas), renewable forms of energy (i.e., hydro, wind, solar, biomass, and geothermal), and nuclear energy, in that order. The impact of the global warming induced by the CO2 emissions from fossil energy conversion processes has become an issue of international concern. An energy solution prompted by the combination of ever-increasing energy consumption and rising environmental concerns thus requires a consideration of coupling fossil energy conversion systems with economical capture, transportation, and safe sequestration schemes for CO2. A long-term energy strategy for low or zero carbon emission technologies would also include nuclear energy and renewable energy. Nuclear power is capable of generating electricity at a cost comparable to the electricity generated from fossil fuels.2 A variety of social and political issues as well as operational safety and permanent waste disposal concerns, however, could limit nuclear energy’s widespread utilization in overall energy * To whom correspondence should be addressed. E-mail: fan@ chbmeng.ohio-state.edu.
production.1,2 Renewable energy sources, although attractive from the environmental viewpoint, face complex constraints for large-scale application. Even when both the decrease in renewable energy costs and the increase in fossil fuel prices are taken into account, it is projected that only about 13.3% of the total energy consumption in 2030 will be from renewable sources. For primarily economic reasons, fossil fuels, including crude oil, natural gas, and coal, will continue to play a dominant role in the world’s energy supply for the foreseeable future. The traditional carbonaceous fuel conversion technologies for combustion or gasification generate flue gas or syngas from which the separation of carbon dioxide requires an elaborative procedure. In contrast, the chemical looping concept allows a sequestration-ready CO2 stream to be directly generated through a different combustion or gasification path. This path is achieved using oxygen carrying particles as the looping media. In this article, the historical and current carbonaceous fuel conversion technologies based on chemical looping concepts are presented. The looping media employed in the processes are mainly in solid form while the carbonaceous fuels can be in solid, liquid, or gas form. Since the success of the chemical looping technology depends strongly on the performance of the chemical looping particles, the subjects of particle properties and selection are discussed. The looping processes can be applied for combustion and/or gasification of carbon-based material such as coal, natural gas, petroleum coke, and biomass directly or indirectly for steam, syngas, hydrogen, chemicals, electricity, and/or liquid fuels production. The test results obtained from the
10.1021/ie1005542 2010 American Chemical Society Published on Web 07/30/2010
Ind. Eng. Chem. Res., Vol. 49, No. 21, 2010
10201
small pilot scale chemical looping combustion and gasification units are described. The various process configurations and energy conversion efficiencies of the chemical looping processes for combustion and gasification applications are given. Novel chemical looping applications, including those for steam-methane reforming and power generation using fuel cell, are also discussed. It is noted that this paper focuses mainly on the metal-metal oxide particle based chemical looping processes in which the chemical looping particle acts as an oxygen carrier. The metal oxide-metal carbonate looping processes,3 which capture carbon through the carbonation reaction of a metal oxide based CO2 sorbent, are not within the scope of this article. 2. History of Chemical Looping and Chemical Looping Based Exergy Optimization Strategy A given reaction can be decomposed into multiple subreactions in a reaction scheme using chemical intermediates that are reacted and regenerated through the progress of the subreactions. A reaction scheme of this nature is referred to as chemical looping. Two types of chemical looping systems, that is, chemical looping gasification (CLG) and chemical looping combustion (CLC), are often used for fossil fuel conversions. The CLG process produces hydrogen from carbonaceous fuels. In most cases, heat is also produced. The CLC process indirectly combusts the fuel to generate heat; no hydrogen is generated in the CLC process. The principles of chemical looping for carbonaceous fuel conversion were first applied for industrial practice between the late nineteenth century and the early twentieth century. Howard Lane from England was among the first researchers/engineers who conceived and successfully commercialized the steam-iron process for hydrogen production using the chemical looping gasification principle.4 With the aid of the iron oxide chemical intermediate, the steam-iron process generates H2 from reducing gas obtained from coal and steam through an indirect reaction scheme: Fe3O4 + 4CO (or H2) f 3Fe + 4CO2 (or H2O) 3Fe + 4H2O f 4H2 + Fe3O4 Net reaction: CO + H2O f CO2 + H2 Looping medium: Fe3O4 T Fe The first commercial steam-iron process, based on the Howard Lane design, was constructed in 1904. Hydrogen plants based on the same process were then constructed throughout Europe and the U.S., producing 850 million cubic feet of hydrogen annually by 1913.5 The steam-iron process only partially converts the reducing gas. Moreover, the iron based looping medium exhibited poor recyclability, especially in the presence of sulfur.5-8 With the introduction of less costly hydrogen production techniques using oil and natural gas as feedstock in the 1940s, the steam-iron process became less competitive and was phased out. In the 1950s, the chemical looping combustion (CLC) scheme was proposed for CO2 generation.9,10 A schematic flow diagram of the fundamental CLC scheme is given in Figure 1. Following Figure 1, the Lewis and Gilliland Process used two fluidized bed reactors, the CO2 generator or reducer and the metal oxide regenerator or oxidizer, for continuous CO2 production. Oxides of copper or iron were used as the looping
Figure 1. Schematic flow diagram of the CLC process.
Figure 2. Exergy recovering scheme for carbon gasification/water gas shift process, scheme I (top) and a simplified chemical gasification process scheme, scheme II (bottom; the reaction temperature is assumed to be 1123 K for scheme II).
particles and carbonaceous fuels such as coal, carbon monoxide, and syngas were used as the feedstock.9 The reaction scheme, which is somewhat similar to that in the steam-iron process, is given below: MeO (oxide of copper or iron) + CO/H2 f Me + CO2/H2O Me + 1/2O2 f MeO Net reaction: CO/H2 + 1/2O2 f CO2/H2O Looping medium: Oxides of copper or iron In the early years, the adoption of a chemical looping strategy was mainly prompted by the lack of effective chemical conversion/separation techniques for product generation. The lack of understanding in oxygen carrier particles development, looping reactor and process design, and chemical reaction engineering renders the early chemical looping process far from optimized. The modern applications of chemical looping processes are prompted by the need for developing an optimized fossil fuel conversion scheme that minimizes the exergy loss involved in energy conversion and CO2 capture, thereby yielding an overall efficient and economical process system.11-17 A simplified exergy analysis given in Figure 2 illustrates the potential application of the chemical looping gasification concept
10202
Ind. Eng. Chem. Res., Vol. 49, No. 21, 2010
for efficient carbonaceous fuel conversion. The analysis involves the following key assumptions and definitions: (1) The environmental temperature is T0 ) 273.15 K and the environmental pressure is P0 ) 101,325 Pa. (2) The reference substance for C and CO is 400 ppm CO2, the reference substance for H2 is water, and reference substance for Fe, FeO, and Fe3O4 is Fe2O3. (3) Coal is considered as pure carbon. (4) Heat can be integrated with a 100% efficiency whenever feasible. (5) The enthalpy of devaluation is defined as the enthalpy change for a substance from the current state to its reference state. (6) The feedstock outside of the reaction loop is at the environmental state. Scheme I represents a traditional route where carbon is converted into syngas followed by the water gas shift reaction to produce hydrogen. As can be seen, the gasification step leads to at least 12% exergy loss because of partial oxidation. The water gas shift reaction after the gasification step leads to another 8.8% exergy loss because of the conversion of CO into H2 that has both lower enthalpy of devaluation and a lower exergy rate. The exergy loss in separating hydrogen from CO2 is not taken into account. Scheme II illustrates an alternative chemical looping gasification approach. Here, Fe is used as a chemical intermediate to convert carbon into hydrogen. There are also two steps involved in scheme II: Step 1
C + 0.395Fe3O4 + 0.21O2 f 1.185Fe + CO2
Step 2
3Fe + 4H2O f Fe3O4 + 4H2
As can be seen, much less exergy loss occurs in step 1 (23.1 kJ) compared to the traditional gasification step (48.8 kJ). Moreover, a zero exergy loss is achievable in step 2 of scheme II through integration of a small amount of low grade heat. As a result, the exergy loss per mole of H2 produced is reduced by more than four times using the chemical looping approach. The results presented in the examples are based on a set of simplified assumptions. Thus, they represent an upper bound to the energy conversion efficiencies in the processes. Nevertheless, the results provide a guide for comparisons of the conventional option versus the chemical looping gasification option. It can also be shown, through simple exergy analysis, that the chemical looping combustion (CLC) process can be more exegetically efficient than conventional power generation processes. To generalize, the exergy loss can be lowered through a well conceived chemical looping scheme comprising steps that are less irreversible. The irreversibility of the looping steps is affected, to a large degree, by the type and performance of the chemical looping medium or particle, the chemical looping reactions, and the chemical looping reactor and system design. 3. Chemical Looping Particles A successful operation of chemical looping processes strongly depends on the effective performance of the chemical looping particles or oxygen carrier particles. Many factors are involved in the design of particles that possess desirable properties. This section discusses the desired characteristics for the chemical looping particles and the role of thermodynamic analysis on the selection of the particle and reactor design. 3.1. General Particle Characteristics. The chemical looping particles are usually composed of primary metal oxide, support,
and promoter or doping agents. The design and preparation of these composite particles involve, to date, a complex trial and error procedure. Possible variables include types of metal oxide and support, weight percentage, and synthesis method and procedure. Effective particle development requires comprehensive consideration of various desired properties of the looping particles including: • Suitable thermodynamic properties. • Good oxygen carrying capacity. • Good gas conversions in both the reduction and oxidation reactions. • High rates of reaction. • Satisfactory long-term recyclability and durability. • Good mechanical strength. • Suitable heat capacity and high melting points. • Ability to change the heat of reaction. • Low cost and ease in scale-up of synthesis procedure. • Suitable particle size. • Resistance to contaminants and inhibition of carbon formations. • Minimal health and environmental impacts. Over 700 different particles have been tested with the focus on redox reactivity, recyclability, and attrition behavior related to chemical looping combustion applications.18,19 Several promising particles have been identified. In terms of primary metal/metal oxides, they are Ni/NiO, Cu/CuO, Fe/FeO, Fe3O4/ Fe2O3, and MnO/Mn3O4 with the support of Al2O3, TiO2, Ni-, Co-, or Mg-Al2O4, bentonite, or ZrO2. The reactivities of various metal/metal oxides with different inert support in reduction reactions with CH4 and oxidation reactions with air at different temperatures have been reported.16 It is concluded that, in general, Ni and Cu have high activities. However, Cu sinters at 950 °C, which limits its application at high temperatures. Fe exhibits a moderate activity with the Fe2O3-Fe conversion usually being incomplete.20 The Mn activity varies with the active metal oxide content and type of inert support. With the ever expanding database on chemical looping particle development, a rational particle design procedure should be developed with consideration of the various physical and chemical characteristics of different primary metal oxide and support, previous testing results, and the specific looping process requirements. A thorough thermodynamic analysis is almost always required for the design of optimum chemical looping particle and reactor system. 3.2. Thermodynamic Properties of Metal and Metal Oxides and Reactor Design Optimization. The capability to capture CO2 is required for modern chemical looping processes. Thus, it would be desired that the primary metal oxide in the chemical looping particle can fully convert carbonaceous fuels into CO2 and H2O while the reduced metal can be fully regenerated back to metal oxide through reaction with the oxidizing agents such as steam, air, or CO2. The maximum extents of the metal oxide reduction and oxidation reactions are dictated by the thermodynamic relationship. Figure 3 shows the thermodynamic phase diagrams of the Ni-H-O and Ni-C-O systems (Figure 3a) and Ni-O system (Figure 3b). As can be seen in Figure 3a, a mixture of NiO and Ni equilibrates with a steam and hydrogen mixture that contains ∼99% steam at 900 °C. This phase diagram indicates that, when pure hydrogen is used as the fuel, the presence of excess NiO can lead to a maximum hydrogen conversion of ∼99% at 900 °C. Similarly, the maximum CO conversion is also ∼99% at 900 °C. Therefore, when used in the chemical looping reducer, NiO can convert nearly all the syngas fuel into a concentrated
Ind. Eng. Chem. Res., Vol. 49, No. 21, 2010
10203
Figure 3. (a) Equilibrium phase diagram for Ni-NiO system for redox reactions with CO2/CO (solid line) and H2O/H2 (dashed line). (b) Equilibrium phase diagram for Ni-NiO and O2.
CO2 and H2O mixture. The analysis on phase diagrams of nickel with other fuels such as methane indicates that NiO is also suitable for oxidizing other carbonaceous fuels. The equilibrium phase diagram of Ni-H-O system shown in Figure 3a also reveals that Ni cannot convert a significant portion of steam into hydrogen. Therefore, it is not suitable for chemical looping gasification applications. Since Ni equilibrates with extremely low partial pressure of oxygen as shown in Figure 3b, it can be easily regenerated to NiO with air in the CLC oxidizer. The above analysis indicates that Ni/NiO particle is suitable to be used for chemical looping combustion; however, it is not suitable to be used for chemical looping gasification. The thermodynamic analysis on metals with two oxidation states such as nickel can be conducted with ease. The thermodynamic analysis on metals with more than two oxidation states such as Fe and Mn, however, is complex since different metal oxides may exhibit significantly different thermodynamic properties. The information from the thermodynamic analysis not only can reveal the suitability of certain types of metal for chemical looping reaction applications, it can also guide the optimum strategy in designing the reducer or oxidizer reactors for the chemical looping reaction operation. To illustrate the relationship between the thermodynamic properties of the metal and metal oxide(s) and the reactor design, two types of oxygen carriers are considered, that is, NiO and Fe2O3. With NiO as the looping particle in the reducer, the maximum fuel and NiO conversions can be achieved irrespective to the type of the multiphase flow reactor used from the thermodynamic standpoint. This is because the oxygen carrier has only two oxidation states, that is, Ni and NiO. With Fe2O3 as the looping particle in the reducer, on the other hand, both the extent of the fuel and the looping particle conversions would differ depending on the types of the multiphase reactors used. This is because the oxygen carrier, Fe2O3, has multiple oxidation states as shown in Figure 4. When a fluidized bed is used as the reducer, the partial pressure ratio of CO2 to CO is expected to be high throughout the reactor. Therefore, Fe2O3 can only be reduced to Fe3O4 as indicated in Figure 4. Thus, using a fluidized bed, the extent of the conversion of Fe2O3 in the reduction reaction is to be low. In contrast, when a countercurrent moving bed reactor is used, the extent of the conversion of Fe2O3 can be high because of a low partial pressure ratio of CO2 to CO at the solids outlet, yielding a reduced product in the form of Fe/FeO. Thus, to achieve a given extent of the fuel
Figure 4. Thermodynamics of iron oxides and CO (solid line)/H2 (dashed line) reactions in fluidized bed and countercurrent moving bed reactors.
conversions in a reducer, a significantly smaller amount of Fe2O3 particles is needed using a moving bed reactor as compared to a fluidized bed reactor. Similarly, a countercurrent moving bed reactor enhances the steam to hydrogen conversion during the regeneration of the reduced Fe/FeO particles.21 Since thermodynamic analysis reveals the intrinsic physical chemical properties of the particle, it could be used as a screening step in selecting the type of looping particles to be used and the type of reactor to be designed. The development of an ideal chemical looping particle requires also to consider such intertwining factors as the reaction kinetics, cost, recyclability, physical strength, ease in heat integration, resistance to contaminants, and environmental and health effects. For metal oxides such as Fe2O3, a countercurrent fuel-particle contact mode conducted in a moving bed can achieve a maximum metal oxide reduction conversion, which is not practically achievable by a fluidized bed. Although there is no omni particle that possesses all the desired properties, as described above, and can be utilized for all looping process applications, a proper selection of a suitable primary metal oxide material, as well as its support and doping agent based on the type of the fuel to be converted and the intended product to be generated, is key to the ultimate development of a suitable
10204
Ind. Eng. Chem. Res., Vol. 49, No. 21, 2010
Table 1. Comparisons of the Key Properties of Different Metal Oxide Candidates20,22-24
a
cost oxygen capacityb (wt %) thermodynamicsc kinetics/reactivityd melting points strength environmental and health effects
Fe2O3
NiO
CuO
CoO
+ 30 + + + ∼
21 + + ∼ -
20 + + ∼ -
21 + + ∼ -
a +, positive; -, negative; ∼, neutral. b Maximum possible oxygen carrying capacity by weight. c Capability to fully oxidize C, CO, and H2 to CO2 and H2O according to thermodynamic principles. d Reactivity refers to the rates of the reactions between metal oxides and syngas (CO and H2).
looping particle. Moreover, optimized reactor design is required to render the process operation economically feasible. The following section illustrates the primary metal selection for one specific application. 3.3. Selection of Primary Metal for Chemical Looping Combustion of Coal. A survey of several metal oxide candidates for chemical looping combustion of coal is given in Table 1. Among the metal oxides, iron oxide is of a low cost and has a high oxygen carrying capacity, favorable thermodynamics, a high melting point, and good mechanical strength. Further, iron induces low health and environmental effects. The reactivity of iron particle is, however, relatively low. NiO is often considered as a good oxygen carrier as it reacts faster with CO or H225 when compared to Fe2O3. However, it is noted that, in the presence of H2O and/or CO2, the reaction rate between metal oxides and coal char is controlled by char gasification rather than the slow solid-solid reaction between coal char and metal oxide:26 H2O/CO2 + C f CO + H2 /CO
(1)
MeO + H2 /CO f Me + H2O/CO2
(2)
CO2 or H2O acts as an enhancer which improves the solid-solid reaction rate. For both NiO and Fe2O3, Reaction 2 is much faster than Reaction 1. Therefore, in the presence of CO2/H2O, the reaction rate between NiO and coal char should be similar to that between Fe2O3 and coal char. Thus, the reactivity advantage of NiO with syngas is “silenced” when using coal in the chemical looping combustion system. Noting that the iron based oxygen carrier particle is far less costly than the nickel based oxygen carriers; thus, iron is a more favorable primary metal for chemical looping combustion of coal. 4. Chemical Looping Combustion Systems As discussed in the introduction, the CLC concept can be traced back to the pioneering study of Lewis and Gilliland in the 1950s when they proposed to use redox reaction of metal oxides to produce carbon dioxide from syngas. In their process, carbon dioxide was the desired product. In the late 1960s, the CLC process was proposed as a novel fuel conversion route that reduces the irreversibility of fuel combustion for heat and power generation.11,14 Verified to be advantageous by thermodynamic analysis especially under a carbon constrained scenario,14,17,27 the chemical looping concept has been extensively explored during the last two decades. Studies carried out in the 1980s and 1990s focused on the development of the chemical looping particles and their applications in CLC processes using gaseous fuels such as methane and syngas.28-31 From the beginning of this century, the possibilities of utilizing
Figure 5. Schematic diagram of the dual circulating fluidized bed (DCFB) system constructed at the Vienna University of Technology.
solid fuels such as coal and biomass in a CLC system have also been considered.32-36 This section discusses the CLC reactor design and operational results. 4.1. Gaseous Fuel CLC System and Operational Results. The promising results obtained from the lab and bench scale experiments in the early 1990s prompted efforts for testing the CLC process at larger scales. Prior operational experiences in industrial circulating fluidized bed (CFB) processes for FCC processing and coal combustion provide a logical basis and a fundamental framework for their extension to CLC applications. As a result, nearly all the existing subpilot scale CLC systems employ a CFB design. The existing subpilot scale CLC testing units include the Chalmers University 10 kWth CLC system,37 the Instituto de Carboquimica (CSIC) 10 kWth CLC system,38 the Korea Institute of Energy Research (KIER) 50 kWth CLC system,39 and the Vienna University of Technology (VUT) 120 kWth CLC system.40,41 This section focuses on the design and operational results obtained from the VUT 120 kWth CLC system. As illustrated in Figure 5, the VUT CLC unit is a dual circulating fluidized bed (DCFB). The DCFB system has a designed fuel power of 120 kWth.40,41 As can be seen from Figure 5, the DCFB system consists of mainly a reducer (fuel reactor), an oxidizer (air reactor), an upper loop seal, a lower loop seal, an internal loop seal, and two cyclones. The system forms two particle circulation loops, the global particle circulation loop, and the local particle circulation loop. The cyclone and internal loop seal placed around the reducer allow for local particles circulation within the reducer, independent of the global particles circulation between the oxidizer and the reducer. The reducer is operated at near the turbulent regime, and the oxidi-
Ind. Eng. Chem. Res., Vol. 49, No. 21, 2010 Table 2. Operational Results of the VUT DCFB Unit (Data with Air/Fuel Ratio
