Joint UK / China Hydrogen production network
Lead Research Organisation:
University of Cambridge
Department Name: Engineering
Abstract
The increasing threat posed by enhanced global warming, together with the requirement to secure energy supplies for both countries have led to this proposal for a collaboration of experts between China and the U.K. in clean technologies for energy production from fossil fuels. The overarching theme of the proposal is the production of clean energy/H2 from coal, via a number of thermochemical routes with the CO2 separated and ready for sequestration. We will investigate two forms of advanced chemical cycles which allow clean hydrogen to be produced from fossil fuels, without (unlike with current technology) a large energy penalty associated with capturing the CO2. These processes are at an early stage of development with research required on the underlying science of the concepts, as well as how these processes can be scaled up from the laboratory. Both types of chemical cycle make use of solid reactants which either act as CO2 acceptors or oxygen carriers. 1. Advanced gasification processes using the calcination-carbonation cycle: The original ZECA process aimed to generate hydrogen from coal, by first hydrogasifying the coal to methane, then reforming to syngas, before shifting to H2. The shift reaction was to be performed using calcium oxide to remove the CO2 and move the equilibrium of the water-gas shift reaction over to H2. The need to first reform the methane in to syngas faced potential problems with the sulphur in the coal which will contaminate the methane as H2S, and deactivate reforming catalysts. Here we will investigate combining the reformer and shift reactors, and the effect of H2S on the calcium looping agent, which must be repeatedly cycled between CaO and CaCO3. The hydrogasification of a spectrum of fuels needs to be explored, since the efficiency of this process will depend on the ability to completely convert the solid fuel into methane. At a pilot scale the continuous operation of enhanced water-gas shift process will be investigated, in a circulating fluidised bed.2. Hydrogen production using the iron-oxide based redox cycle: In chemical looping combustion, a fuel can be burned with a metal oxide (rather than air) to produce a stream of pure CO2. For power generation, the reduced oxide can be reoxidised with air to release heat. Some metals and oxides (e.g. iron) can be partially oxidised with steam to produce very pure H2. Fe2O3 oxide can be reduced to FeO or Fe using syngas; Fe and FeO can then be oxidised with steam giving Fe3O4 and hydrogen. The cycle can be completed by oxidising the Fe3O4 with air. Here we will investigate the continuous operation of this process on a laboratory scale, and on a pilot scale, using a combination of fluidised and moving bed reactors. The syngas must be generated from coal and will contain tars and H2S. We will investigate the affect of volatile material and sulphur on the iron based carrier, i.e. the extent to which the metal oxides can combust the volatiles, and whether the oxides are deactivated by sulphur. The use of these metal oxides ad tar cracking catalysts during gasification will also be investigated. Both the calcium based CO2 acceptors and the metal oxide based oxygen carriers must undergo many cycles of operation. Natural materials will often rapidly degrade. Artificial particles can be produced which have better characteristics. However, the behaviour of the particles is a very strong function of the physical structure, and the presence of additives/contaminants. We will investigate how the formulation of these materials affects their physical structures and the impact this will have on the reactivity over many cycles.
Organisations
Publications
Bao J
(2013)
Interaction between Fe-based oxygen carriers and n-heptane during chemical looping combustion
in Proceedings of the Combustion Institute
Dunstan M
(2013)
Reversible CO 2 Absorption by the 6H Perovskite Ba 4 Sb 2 O 9
in Chemistry of Materials
Jason Cleeton (Author)
(2011)
Interaction between Fe-based oxygen carriers and volatile hydrocarbons during Chemical Looping
Kidambi P
(2011)
Interaction of Iron Oxide with Alumina in a Composite Oxygen Carrier during the Production of Hydrogen by Chemical Looping
in Energy & Fuels
Liu W
(2012)
The Effect of Addition of ZrO 2 to Fe 2 O 3 for Hydrogen Production by Chemical Looping
in Industrial & Engineering Chemistry Research
Liu W
(2014)
Kinetics of the reduction of wüstite by hydrogen and carbon monoxide for the chemical looping production of hydrogen
in Chemical Engineering Science
Liu W
(2015)
Inhibiting the interaction between FeO and Al 2 O 3 during chemical looping production of hydrogen
in RSC Advances
Liu W
(2014)
Performance of a Fe2O3-based oxygen carrier, stabilised with NaAlO2, for the chemical looping production of hydrogen
in Digital proceedings of the 3rd International Conference on Chemical Looping
Liu W
(2012)
An investigation of the kinetics of CO2 uptake by a synthetic calcium based sorbent
in Chemical Engineering Science
| Description | Summary: In this grant we examined the use of chemical looping as a means of producing pure H2 from coal. In this process the CO2 can be captured as part of a Carbon Capture and Storage (CCS) scheme; i.e. allowing a carbon free fuel to be produced from coal. In this process the coal is gasified to syngas, and an oxygen carrier (such as hematite, an iron oxide) is reduced using the syngas (e.g. to iron). Hydrogen is produced be reacting the iron with steam to produce magnetite (an iron oxide), before returning the oxide to its original state by oxidising with air. This process is also known as the steam-iron process. Thus, as a cyclic process the oxygen carrier must be able to undergo many cycles without degrading and losing activity. Typically pure iron oxide will lose all reactive when reduced to metallic iron in only a few cycles. We have developed several new materials which can survive many cycles of oxidation and reduction by combining the active material (iron oxide) with various support materials. Importantly, detailed characterisation of the oxygen carrier materials has given insights into why these materials have a higher performance than the unsupported materials, opening opportunities to design new oxygen carriers. In some case, the support materials (e.g. alumina) were playing an active role in the cycles, by forming solid solutions and other compounds (i.e. Iron Aluminate) at various points in the redox cycles. Other materials proved to be more inert (e.g. ZrO2) with not large scale interaction with the active component. In the case of the ZrO2 support, the support may have increased the accessibility of the iron as the material began to sinter. From our work it is also clear that a detailed understanding of the material chemistry is needed to understand the oxygen carrier behaviour, for example the high performance of one particular oxygen carrier could be linked to the presence of sodium introduced as a contaminant during manufacture, changing how the active oxide (in this case copper) and the support material interacted. We have developed detailed kinetic models for various aspects of the hydrogen production process including for gasification of the char and reduction of the oxygen carriers. The enhancement to gasification which is seen in the presence of some chemical looping materials (e.g. in direct solid fuel chemical looping) was investigated in detail; this important since in many gasification based technologies the conversion of the coal char is difficult. Through a combination of experiment and detailed modelling we were able to explain this enhancement and suggest reactor conditions which might exploit it to best effect. Coal is a complex solid fuel, with a large part of its heating value held in "volatile" material which the coal releases at it is heated. Thus, in any process using coal the interaction of the volatile material and that of the resulting carbon which remains (i.e. the char) must be investigated. Volatile material is responsible for the formation of tars and often causes problems for down-stream equipment. Our investigations show that the active oxygen carrier are very good at destroying most hydrocarbon species, perhaps with the exception of some of the more stable compounds with ring structures. We also observed much cleaner off gases when gasifiying in a bed of oxygen carrier. Thus, most volatile species are unlikely to pose problems in a chemical looping system, since they are much more easily combusted than the char. Some individual findings from our work are detailed below: We carried out an extensive study of the kinetics of gasification of different rank coals in both an inert fluidised bed (e.g. of sand), a bed of depleted oxygen carrier and a bed of active oxygen carrier (i.e. able to give up oxygen). We deliberately excluded oxygen carriers which would release gas phase oxygen so as not to introduce additional combustion reactions. We were able to produce an accurate model for the rate of gasification of the lignite char (which proved to be most appropriate for use in a chemical looping process owing to the high reactivity of the char) and use this in later work to isolate the mechanism by which chemical looping accelerates the rate of gasification of the char. This model was not only able to account for the initial rate under chemical reaction control, but also the transition to internal and external mass transfer controlled kinetics. The variation of reactivity with conversion was also captured using a novel model free approach. The cylindrical pore model was applied for the first time to char gasification to account for intra particle effects. This work is detailed in a paper in Fuel which is currently awaiting publication. Large differences between the rate of gasification of a char in an inert bed and in a bed of oxygen carrier were found. Interestingly, the depleted oxygen carrier had little effect on the rate of gasification, indicating that the presence of metals in the bed alone isn't sufficient to accelerate the rate. Although no effect was anticipated, questions are always raised about the potential for metal to migrate from the fluidised bed particles to the char and catalyse the gasification reaction. We further eliminated this effect by deliberately dosing the chars with various levels of metals to see the effect of metal ion contamination on the rate of gasification. Although this produced some effect, it was too small to account for the changes in reactivity seen when an active oxygen carrier was present. The most noticeable observation was that the oxygen carrier caused the largest increase in reactivity for the highest reactivity chars, and at their peak gasification rate (at > 20% conversion of the char). Order of magnitude calculations also indicated that the char gasification rate was being affected by external mass transfer in these circumstances. Although it is known that gasification is inhibited by the products of gasification (H2 and CO), and that active chemical looping agents will convert these to H2O and CO2, this affect was not sufficient to explain the rate increases which were observed. This led us to propose a mechanism by which the conversion of the products of gasification in the boundary layer by the chemical looping agent led to (1) a drop in inhibitory products at the surface of the char particle (2) a local production of reagent for gasification at the surface of the char particle (3) an increase in the effective rate of mass transfer to and from the particle. These findings were presented at the 4th IEAGHG meeting on high temperature looping cycles. Using detailed models for the gasification of the char we developed earlier in the grant, and measured kinetics for an iron based chemical looping oxygen carrier, we were able to produce a model which fully explained the enhancement to gasification rate seen when a char is gasified in a bed of oxygen carrier. This confirmed that our original hypothesis that the looping agent destroys H2 and CO in the boundary layer surrounding the char and amongst other effects, enhances the rates of mass transfer to and from the particle of char. A complete description of reaction + multi-component diffusion within the porous gasifying char particle, and outside in the boundary layer surrounding the particle was included in the model. The completed model was able to match experimental results and, notably, contained no new chemistry or physics, i.e. the enhancement of rate arose from synergistic effects between already well know mechanisms. Another noticeable result was that enhancement of gasification was most pronounced when the char had undergone some conversion so that the pore structure had opened up, altering the balance between internal and external mass transfer effects. With our model we were able to suggest operating conditions and regimes where significant enhancement of rate would be expected. This work is detailed in a paper which is currently awaiting publication. Owing to the fact that low reactivity solid fuels produce chars which are difficult to gasify, some work was carried out with materials which exhibit a "CLOU" effect, i.e. they can release gas phase oxygen allowing partial combustion to take place. We were able to manufacture copper-alumina based materials with a relatively well controlled microstructure by precipitating Layered Double Hydroxide (LDH) structures, which on calcination gave a plate like structure. The oxygen carriers produced in this way were exceptionally stable, even in extreme conditions, with the platelet like structure (which gives a macro-porosity) surviving in many parts of the oxygen carrier even when cycled at high temperatures. The oxygen release capacity remained constant over many cycles of oxygen release and oxidation. One surprising results, which perhaps explains some of the variability seen in previous work, is that the oxygen carriers which worked best were contaminated by sodium which precipitated out during the manufacture. The sodium prevented the formation of copper aluminate and kept the alumina "support" in the amorphous state (whereas equilibrium would predict crystallisation to alpha alumina or reaction with the copper oxide to form copper aluminate). Thus, rather than the alumina forming copper aluminate and taking part in the cycling reactions, it remained largely inert. This inertness of the support, and the good initial dispersion probably accounted for the stability of the structures seen in the particles of oxygen carrier. These findings are documented in detail in a paper in Energy and Environmental Science (http://dx.doi.org/10.1039/C2EE22801G). These materials were also tested with coal and were found to be very effective in combusting the char and volatile matter. Extensive tests were carried out on the interaction between volatile material and the iron based oxygen carriers (either pure iron-oxide or iron oxide doped with alumina to improve the long term cycling performance) used in the experiments with coal. Some experiments were carried out with various model volatile species, e.g. toluene, heptane and iso-octane in a TGA. When fuels are introduced into a TGA, thermal decomposition of the fuel can take place, meaning that the oxygen carrier is interacting with a complex mixture of species. The oxygen carrier did not affect the temperature at which thermal decomposition of the fuel took place, but showed high reactivity towards the decomposition products of n-heptane and iso-octane. Poor reactivity towards toluene was attributed to the relative difficulty of cracking toluene into species which did not contain very stable benzene rings. With pure iron oxide there was little carbon accumulation until the oxygen carrier was almost completely reduced, perhaps indicating that deposited carbon can be combusted by the oxygen available in the solid. Adding Al2O3 to the oxygen carrier improved the stability of oxygen release over many cycles, but resulted in more carbon deposition, perhaps because of the formation of iron aluminate which would reduce the amount of oxygen which is readily available. This research shows that Fe-based oxygen carriers can combust many of the hydrocarbon species, which may be released during gasification of coal or biomass, to CO2 and H2O. Some species may prove difficult to combust, e.g. those containing benzene rings. Adding Al2O3 to the oxygen carrier may cause problems with carbon deposition, and hence result in H2 which is contaminated with carbon oxides in a steam-iron hydrogen production system. Details of this research were presented at the Clearwater clean coal conference in 2011. The fate of hydrocarbon species in a fluidised bed of oxygen carrier was studied in a fluidised bed by feeding a model volatile compound (n-heptane) to a bed of iron-oxide based carrier. Feeding n-heptane to an inert bed led to the formation of acetylene, followed by (at high temperatures) significant carbon deposition. Less carbon was deposited on the Fe90Al carrier (containing 90 mol% Fe2O3 with 10% alumina). Experiments in the TGA showed significant carbon deposition began when the pure iron oxide carrier was reduced to metallic Fe. However for the Fe90Al material, coking began when the samples was reduced to a mixture of Fe + FeAl2O4 and the reduction of FeAl2O4 had started. A fluidised bed of pure Fe2O3 carrier agglomerated when the carrier was reduced beyond wustite, whilst in contrast the Fe90Al carrier did not agglomerate even when reduced to Fe and FeAl2O4 and remained active despite sintering of the microstructure. Similar observations were made when using syngas as the reductant. This research was carried out in collaboration with partners from Tsinghua University during a visit to Cambridge and led to a joint publication in the Proceedings of the Combustion Institute (http://dx.doi.org/10.1016/j.proci.2012.07.079) and a presentation at the 34th Combustion Symposium. Iron oxide carriers which contained alumina were heavily used in this grant. Previous work had shown that adding alumina improved the ability of the oxygen carrier to be cyclically reduced and re-oxidised in a steam iron process to make H2. Some work was carried out to understand the role of alumina in these systems. Mixing the active material (iron oxide - Fe2O3) with a supposed inert material was thought to prevent sintering and agglomeration by "diluting" the active material. Our work showed that the alumina cannot be considered inert, and does participate in the reaction by forming solid solutions with the various iron oxides and also iron aluminate, FeAl2O4 (also known as hercynite). Thus, we studied the use of solid solutions of _-Fe2O3 and _-Al2O3, synthesized by the co-precipitation method as oxygen carriers in chemical looping for the production of hydrogen. When the carrier contains alumina the kinetics of the reactions and final equilibrium conversion are altered. When alumina is present the fact that solid solution can form means that the equilibrium conversion can be a smooth function of the oxygen fugacity in the gas phase; In contrast, pure iron oxide gives step changes at fixed oxygen fugacities corresponding to the Fe2O3/Fe3O4, Fe3O4/FeO and FeO/Fe transitions. The fact that hercynite forms on reduction also may limit the extent of conversion of the iron oxide to Fe, since the hercynite that forms is more difficult to reduce than wustite. The behaviour of the oxygen carriers was found to correspond reasonably well to phase diagrams for the Fe-Al-O system available from the geology/materials literature. When the particles of oxygen carrier were subjected to up to 50 cycles of reduction and oxidation in a fixed bed reactor at 850 C it was found that the performance began to stabilise after around 20 cycles. Interestingly, the particles had to be fully oxidised using air after each reduction cycle to maintain reactivity, and omitting this step led to a steep decline in performance. The final oxidation in air was found to remove any hercynite that forms and resulted in separate phases of iron oxide and hematite forming; this seemed to correlate with the formation of large macro-pores which allowed reactant gases to penetrate the particle. Thus, whilst cycling destroyed the micro-pores, the formation of macro-pores allowed reactivity to remain high, and the reactivity did not seem to rely on the presence of a large internal surface area. These findings are reported in a paper in Energy and Fuels ( http://pubs.acs.org/doi/abs/10.1021/ef200859d) As a contrast to adding a support material which interacts with the iron oxide during redox cycles, we also investigated other more "inert" supports. Unlike alumina which tends to form aluminate or solid solutions with iron oxide, zirconia (ZrO2) was chosen because of its inertness. The oxygen carriers were prepared by precipitation with various loadings of Fe2O3 and ZrO2. Experiments in the TGA and in a fluidised bed reactor showed that the presence of ZrO2 did not influence the kinetics of reduction of the iron oxide or its final equilibrium state. Further ex-situ characterisation confirmed that there was no significant interaction between the iron oxide and support material. As prepared, the particles contained two phases of ZrO2 (where as thermodynamics for pure ZrO2 would predict only one stable phase), and this did not change upon cycling. Small amounts of dopants are often added to ZrO2 to stabilise its phase (e.g. yttria-stabilised zirconia which is used for oxygen conducting material at high temperatures). Here we proposed that a small amount of Fe3+ did incorporate itself within the zirconia, preventing the metastable tetragonal ZrO2 from changing phase. In addition we found that the oxygen carriers were remarkably stable with cycling, meaning that all the iron oxide remained accessible, even though there was some sintering of the structure. The fact that tetragonal ZrO2, doped with other metallic cations is known to contain a large number of oxygen vacancies, allowing diffusion of oxygen ions through the support material, led us to speculate that gradients in oxygen fugacity in the particle could allow oxygen to be removed from iron oxide which was not accessible to reducing gas. Carbon deposition was found only when the conversion from Fe3O4 to Fe exceeded 90% and there was little degradation in performance of the oxygen carrier over many redox cycles. These results are documented in more detail in a paper in Industrial and Engineering Chemistry Research (http://pubs.acs.org/doi/abs/10.1021/ie302626x) Work on measuring the kinetics of reduction of oxygen carriers in this grant has focussed on the iron based materials, which can used in the steam-iron chemical looping cycle to make hydrogen. Previous work had produced a reasonable kinetic model for the reduction to wustite, but not for the reduction to iron (which can increase the yield of hydrogen produced dramatically). Experiments in the TGA and fluidised bed indicated that the process was initially controlled by nucleation followed by diffusion through a product layer. We were able to produce a detailed mechanism for the reduction of wustite to iron and a kinetic expression which can be used in models of chemical looping processes. The effect of pressure on rate of reduction of the iron based materials was also studied. As expected there was an increase in rate with pressure, and despite difficulties with the speed of the reactions compared with mixing times in the pressurised fluidised reactor, we were able to determine how reduction rate of the carrier will scale with reactor pressure. This initial study on the effect of pressure is timely, as there is now increased interest in pressurised chemical looping combustion. We have successfully designed and commissioned a continuous system for studying the steam iron reactions. The reducing section of this reactor can be operated as either a moving or a fluidised bed. The key challenge was to design a small enough system to be able to run it in the laboratory, but still large enough to allow continuous operation (flowing fluidised systems are difficult to run at small scale). Extensive cold modelling was done to ensure that the reactor could operate, with the result that the final design was the second iteration, the first suffering problems with blockages. This reactor system in addition to being used for the steam-iron cycle is also finding uses in other related EPSRC projects. Whilst most of the work at Cambridge focussed on the metal oxide cycle part of the grant, a small amount of work was carried out looking at the sorbent enhanced shift method for producing H2, in particular modelling the calcium based sorbents which have been developed at Cambridge to capture the CO2 in these processes. These sorbents contain calcium oxide mixed with alumina and are formed by a hydration method. In-situ XRD was performed to track the change in phase during a cycle of carbonation and calcination. This verified that the calcium aluminate (mayenite) which formed when the sorbent was manufactured remained inert. The results of experiments in both a TGA and fluidised bed system, led us to develop a detailed model for the reaction of the sorbent particles, based on the overlapping grain model. This model was found to apply equally well to artificial sorbents described above and natural limestone based sorbents. In the model, the microstructure was described by a distribution of grain sizes, which as an input to the model allowed the same model to be used for different sorbents. Despite the difficulty in obtaining the grain size distribution from measurements made on the sorbents, the model was able to accurately describe the reaction, and account for the superior performance of the artificial sorbents. The model also accounted for changes in rate and extent of reaction during cycles of carbonation and calcination as the structure of the sorbents evolved. These findings are described in detail in a paper published in Chemical Engineering Science (http://dx.doi.org/10.1016/j.ces.2011.11.036). |
| Exploitation Route | The research has industrial applicability. The work has shown that metal oxide cycles can be exploited to produce H2, with a number of materials demonstrated which can do this. More immediate application may lie in smaller scale application where the oxygen carrier materials can be exploited for other purposes, e.g. gas cleaning and selective catalytic oxidation. We have discussed exploitation with a number of companies, and are actively working with an SME on some of these concepts. We have also disseminated our results to various industrial partners. |
| Sectors | Chemicals,Energy,Environment |
| Description | The findings have been used in various academic projects and led to follow on projects (as noted in the follow on funding). |
| Description | Multi-scale evaluation of advanced technologies for capturing the CO2: chemical looping applied to solid fuels. |
| Amount | £578,218 (GBP) |
| Funding ID | EP/I010912/1 |
| Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 04/2011 |
| End | 09/2014 |
| Title | Small Scale laboratory fluidised bed reactor suitable for pressurised operations |
| Description | New reactor system for pressurised work and materials testing. |
| Type Of Technology | New/Improved Technique/Technology |
| Year Produced | 2012 |
| Impact | none |
| Title | Small scale laboratory reactor for continuous chemical looping. |
| Description | We have successfully designed and commissioned a continuous system for studying the steam iron system. The reducing section of this reactor can be operated as either a moving, or fluidised bed. The key challenge was design a small enough system to be able to run it in the laboratory, but still large enough to allow continuous operation (flowing fluidised systems are difficult to run at small scale). This reactor system in addition to be used for steam-iron is also finding uses in other related EPSRC projects. |
| Type Of Technology | New/Improved Technique/Technology |
| Year Produced | 2011 |
| Impact | Further research work carried out with materials developed in other grants. |
| Description | Novel Methods for Hydrogen Production |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | Talks given at IMechE event on the production of Hydrogen, 7th March 2019, London |
| Year(s) Of Engagement Activity | 2019 |