CHEMICAL ENGINEERING TRANSACTIONS VOL. 52, 2016 A publication of The Italian Association of Chemical Engineering Online at www.aidic.it/cet Guest Editors: Petar Sabev Varbanov, Peng-Yen Liew, Jun-Yow Yong, Jiří Jaromír Klemeš, Hon Loong Lam Copyright © 2016, AIDIC Servizi S.r.l., ISBN 978-88-95608-42-6; ISSN 2283-9216 Novel Process for Hydrogen Production Through the Sorption Enhanced Reforming of Methane Combined with Chemical Looping Combustion Jose R. Fernández*, J. Carlos Abanades Spanish Research Council, INCAR-CSIC, C/ Francisco Pintado Fe, 26, 33011, Oviedo, Spain jramon@incar.csic.es A novel process scheme is analysed for the production of hydrogen from methane, while obtaining a separate stream of CO2, which is suitable for a subsequent reuse and/or for its permanent geological storage. This system comprises three interconnected fluidized-bed reactors operating at atmospheric pressure. H2 is produced in a sorption enhanced reforming (SER) stage, where the presence of a CaO-based sorbent allows the CO2 to be removed as it is formed. The main novelty of the process concerns the regeneration of the CaCO3, required to carry out a multicycle operation. A redox chemical loop supplies the heat necessary for the calcination. The CaCO3 formed in the SER is put in contact with a very high-temperature solids stream, which is mainly composed of Fe2O3 resulting from the exothermic oxidation of Fe3O4 with air in a separate reactor (at 950 ºC - 1,200 ºC). The calciner also acts as a fuel reactor, where Fe2O3 is reduced again to Fe3O4 by feeding a fuel gas, which is converted to CO2 and steam. A preliminary thermodynamic assessment of the proposed system has been carried out reaching H2 equivalent efficiencies of up to 79 %, which are around 9 % points above the efficiency of a reference H2 production plant based on conventional steam reforming including CO2 capture with MDEA. Carbon capture efficiencies of around 98 % can be obtained, which are significantly higher than those obtained in the reference plant that uses MDEA absorption (around 85 %). These results confirm the potential of this novel system for a future development as hydrogen production process with CO2 capture. 1. Introduction Hydrogen has long been an important raw material for the manufacture of commodity chemicals such as ammonia and methanol. Increased quantities of H2 are now being used in petroleum refining, metallurgical processes and also in recent applications as fuel cells and combustion engines, which could result in large H 2 demands in a medium-to-long term (IEA, 2006). There is also global concern about reducing anthropogenic CO2 emissions to the atmosphere to mitigate global warming. Among the different options presented, carbon capture and storage arises as a robust and viable solution to drastically reduce global CO2 emissions (IPCC, 2014). Steam Methane Reforming (SMR) is the most widely used technology to produce H2 at commercial scale. SMR is typically carried out in Fired Tubular Reformers (FTR) at high temperatures and high pressures to overcome the thermodynamic constraints of reforming reaction and thus reach high H2 yield. Moreover, SMR requires a complex arrangement composed of high number of stages with different catalysts and consumes large amount of energy due to the high endothermicity of the reforming reaction (Li et al., 2016). In the Sorption Enhanced Reforming (SER), the use of a CaO-based material combined with the reforming catalyst allows the CO2 to be removed as soon as it is formed from the gas phase, generating CaCO3. The equilibrium is then shifted to H2 formation, with the result that almost complete conversion of the fuel gas to H2 is achieved in one single stage under moderate conditions of pressure and temperature. CaO-based sorbents can be obtained from a wide range of abundant, natural and inexpensive sources, such as limestone, dolomite or calcium hydroxide and their presence in the reformer minimizes coke formation and diminishes the steam requirements of the process (Harrison, 2008). The CO2-sorbent needs to be continuously regenerated to carry DOI: 10.3303/CET1652090 Please cite this article as: Fernández J. R., Abanades J. C., 2016, Novel process for hydrogen production through the sorption enhanced reforming of methane combined with chemical looping combustion, Chemical Engineering Transactions, 52, 535-540 DOI:10.3303/CET1652090 535 out a multicycle operation. The calcination of CaCO3 is a highly endothermic reaction that demands temperatures around 900 ºC in pure CO2 and at atmospheric pressure. The main challenge to develop the SER technology at a large scale lies in finding a process scheme that allows the energy required for the calcination to be supplied with minimum energy penalties. In this work, a Fe2O3/Fe3O4 chemical loop is proposed to supply the heat for the calcination on the basis on the scheme described by Fernández and Abanades (2016), which was envisaged for the calcination of CaCO3 in cement plants. The Chemical Looping Combustion (CLC) principle consists in the transfer of oxygen from air to the fuel using a solid oxygen carrier (typically a metal oxide), avoiding any direct contact between the fuel and air. In a first stage (fuel reactor), the fuel is oxidized to CO2 and steam by a metal oxide that is reduced. In a second stage (air reactor), the reduced solid is oxidized with air and the material is regenerated to start a new cycle. The net chemical reaction over the two stages is the same to direct combustion of the fuel with air. Therefore, it is possible to carry out a combustion process with inherent CO2 separation with minimum energy penalties (Adanez et al., 2012). Iron-based oxygen carriers are especially suitable for this application because they are cheap materials and present good stability under repeated oxidation/reduction cycles at high temperatures. This work proposes a new process for H2 production with CO2 capture using a CaO-based sorbent, where the calcination of CaCO3 is carried out by means of indirect heating using iron-based carriers that come at very high temperature from the air reactor of a CLC system. The main operating conditions and design parameters are discussed using the available knowledge in the literature on the key reactions involved. 2. Process description The H2 production process proposed in this work (see Figure 1) comprises three main reaction stages, which are carried out in a compact arrangement of interconnected fluidized-bed reactors operating at atmospheric pressure. In the first stage, a low pressure H2-rich gas (stream 5 in Figure 5) is produced through the sorption enhanced reforming (SER) of CH4 (stream 2) in the presence of a Ni-based reforming catalyst combined with CaO, that acts as CO2 sorbent. This stage is expected to be fast and efficient in a circulating fluidized bed (Reformer Carbonator) at a temperature around 650 ºC, as it is the case in CFB carbonator reactors in calcium looping post-combustion systems (Arias et al., 2013). Under these conditions, both reforming and carbonation reactions are highly favoured, so that almost complete conversion of CH4 to H2 and a drastic CO2 removal from the gas phase can be achieved (Harrison, 2008). The product gas will contain around 90 - 95 % vol. of H2 (depending on the steam-to-carbon ratio in the feed) and less than 1 % vol. of CO and CO2. After cooling and heat recovery, gas from the reformer is compressed to be purified in a pressure swing adsorption (PSA) unit, where high pressure pure H2 (stream 6) is obtained. After the separation from the product gas in a downstream cyclone, the carbonated particles formed in the SER (stream 7) come into a second fluidized bed (Fuel Reactor Calciner) to carry out the regeneration of the CO2 sorbent. The heat required for the calcination of the CaCO3 is supplied by a very high-temperature stream of iron-based oxygen carrier (stream 17), which is mainly composed of Fe2O3 resulting from the exothermic oxidation of Fe3O4 with air in a separate reactor (Air Reactor) at temperatures higher than 950 ºC. During the calcination stage, Fe2O3 is reduced again to Fe3O4 by feeding a fuel gas, which is converted to CO2 and steam. The use of the low pressure PSA off-gas (stream 10), which still contains large fractions of H2, CO and CH4 (Table 1), as reducing agent significantly reduces the need to feed additional CH4 to the Fuel Reactor Calciner, which leads to a higher overall efficiency. The typical temperatures required for the rapid calcination of CaCO3 (860 ºC - 910 ºC) are favourable the reduction of Fe2O3 to Fe3O4 to be effective for a wide range of fuels (Yan et al. 2015). The product gas (stream 11) will be mainly composed of CO2 and steam, but low amounts of unburnt compounds resulting from the incomplete oxidation of the fuel gases can be also present. In this case, an additional oxygen polishing step to com plete the combustion of the gas is required in order to obtain a final concentrated stream of CO2 at the exit of the compression and purification unit (CPU), which is suitable for reuse or purification, compression and storage. In order to reduce the energy demand of the calcination stage, a preheating system composed of at least two cyclones in series is proposed to increase the temperature of the carbonate particles (Romano et al., 2011) before entering the fuel reactor calciner. The solids enter at the top side and while they are passing through the cyclones are heated by a gas stream at a higher temperature that flows in countercurrent. The high- temperature gas (stream 18) generated in the Air Reactor can be used in a first heating stage and the CO2- rich gas obtained in the fuel reactor (stream 11) can be used in the last cyclone to heat the solids up to temperatures around 750 ºC avoiding any previous carbonate decomposition (Fernandez and Abanades, 2016). From the reduction/calcination stage, a mixed stream of solids mainly containing Fe3O4 and CaO is obtained. The use of a downstream segregator will highly improve the performance of the system, since large amounts of iron oxide would act as thermal ballast in the SER stage, while the oxidizing atmosphere and the high temperatures achieved in the Air Reactor would negatively affect the reactivity of the reforming catalyst and the CaO (Gonzalez et al., 2008), respectively, after multiple cycles. 536 Figure 1: Scheme of the proposed H2 production process combining SER and Chemical Looping Combustion The segregator needs to be operated under bubbling-bed regime in order to facilitate the segregation of the Fe-solids from the calcium-based and reforming catalyst before they enter the subsequent reaction stages of the process. The driving force for the separation of the solids will be the ratio of particle densities and of particle sizes between the solids (Naimer at al., 1982). This is a similar operation to that envisaged for carbon strippers under testing in other CLC systems with solid fuels to separate unconverted char from the oxygen carrier (Adanez et al., 2012). Under suitable fluidization conditions, very high separation efficiencies can be achieved (up to 95 %, comparable to those obtained in carbon strippers of CLC), since the CaO and reforming catalyst particles (porous solids with apparent density of around 1,800 kg/m3) will tend to float, while the denser Fe-based particles (with apparent density of around 4,000 kg/m3) will sink in the reactor. The solid stream rich in CaO and reforming catalyst (stream 14) is directed to the Reformer Carbonator for a subsequent SER stage, while the solid stream rich in iron oxide (stream 15) is fed into the air reactor, which operates as a circulating fluidized-bed (CFB), being oxidized with air (stream 16). The oxidation is carried out with a little excess of oxygen to ensure the complete oxidation of the Fe-based particles. The maximum temperature achieved in the Air Reactor will depend on the active content of the Fe-based solids, which will affect the heat transport capacity form the Air Reactor to the Fuel reactor Calciner. The O2-depleted air released at very high temperature (stream 18) is a high quality source of thermal power that can be used for preheating requirements and/or also for power generation in a steam cycle. Meanwhile, the oxidized solids (stream 17) will be fed into the Fuel Reactor Calciner to complete the cyclic operation. 3. Results and discussion Mass and energy balances involved in every stage of the proposed process were solved including a complete thermal integration between the solids and gas streams included in Figure 1. Adiabatic conditions in all the units were assumed taking the thermodynamic properties of the substances involved in the process from Barin (1989). Commercial Fired Tubular Reformer (FTR) plant produces around 30,000 Nm3/h of H2 (about 90 MW) at 29 bar, which is a common size for H2 plants installed in refineries. A reference case assuming this production of H2 in the SER stage of the proposed process is analyzed to theoretically demonstrate its viability. Reference H2 production plant with CO2 capture performed by means of methyldiethanolamine (MDEA) chemical absorption is considered in this analysis for purposes of comparison. A detailed description including a comprehensive thermodynamic assessment of both reference H2 plants with and without CO2 capture can be found elsewhere (Martínez et al., 2014). Different performance indexes have been defined to analyze global operation of the proposed process and compare it with those of the reference plants with and without CO2 capture. Hydrogen yield is defined as moles of H2 produced per mole of CH4 fed into the SER stage. Overall H2 yield is referred to the total moles of CH4 introduced into the process. Hydrogen production efficiency (ηH2) is the ratio between the energy of the H2 output and the energy of the CH4 fuelled into the plant (both based on LHV). To consider the contribution of the electricity and heat flows exchanged with the exterior, an equivalent H2 production efficiency (ηeq,H2) has been defined (Martínez et al., 2014). The CO2 capture efficiency is evaluated using the carbon capture ratio (CCR) calculated as the ratio between the mass flow rate of CO2 arising from the CO2 compression and purification unit (CPU) and the mass flow rate of CO2 associated to the total CH4 fed into the plant. 537 Table 1: Temperature, flow rate and composition of the main gas streams of the proposed process in Figure 1 Streams 1 2 3 4 5 6 10 11 12 13 16 18 19 T, ºC 25 350 350 25 650 35 35 870 742 30 25 975 711 Flow,mol/s Composition, vol.% 131 119 296 13 528 372 73 217 217 126 364 293 293 CH4 100 100 - 100 2.3 - 16.3 1.8 1.8 - - - - H2O - - 100 - 15.8 - - 39.4 39.4 - - - - H2 - - - - 78.3 100 56.7 - - - - - - CO - - - - 2.6 - 18.4 - - - - - - CO2 - - - - 1.2 - 8.7 59.4 59.4 97.0 - - - O2 - - - - - - - - - 2.0 21.0 1.8 1.8 N2 - - - - - - - - - 1.0 79.0 98.2 98.2 In this study, the Ca-based material is assumed to contain 20 wt.% of active CaO and the Fe-based solid consists of 30 wt.% of Fe2O3 over alumina, which are conservative values compared to the usual active contents of these materials in CLC systems (Ge et al., 2016) and calcium looping applications (Martínez et al., 2014). A typical Ni-based reforming catalyst over alumina with an active phase of 10 wt.% is adopted. A catalyst/sorbent ratio present in the Reformer Carbonator of about 0.3 will ensure the operation close to SER equilibrium (Martínez et al., 2014). PSA is assumed to separate 90 % of inlet H2 as high pressure and pure gas (Martínez et al., 2014). Due to the large difference in solids densities, very high separation efficiency (about 95% in lighter solids and 100 % in denser solids) can be expected in the segregator (Fernández and Abanades, 2016). As a result, the solids stream coming to SER stage from the segregator (stream 14) will not contain Fe-based particles, while the solids stream guided to the Air Reactor (stream 15) will only contain less than 2 wt.% of CaO and reforming catalyst. Table 1 indicates the temperature, flow rate and composition of the main gas streams obtained by solving the mass and heat balances at each stage of the proposed process. As can be seen in Figure 2a, the SER operation must be carried out at temperatures no higher than 650 ºC to ensure a product gas with a very high content of H2 (higher than 90 vol.% on a dry basis) and negligible amounts of CO2. High steam-to-carbon molar ratios (S/C) promote the conversion of CH4 to H2 at the expense of higher steam requirements (and therefore higher energy demand). Figure 2: a) Effect of steam-to-carbon molar ratio (S/C) and temperature on sorption enhanced reforming (SER) equilibrium at atmospheric pressure. b) Effect of the S/C molar ratio (using PSA) on the H2 yield Performing the SER stage at 650 ºC with a S/C molar ratio of 2.5 (119 mol/s of inlet CH4 in the reference case), a flow gas of 528 mol/s is produced in the Reformer Carbonator containing around 93 vol.% of H 2, on a dry basis (see Table 1). A flow of solids of 31.6 kg/s (stream 14) comes from the segregator at 870 ºC. Under these conditions, both inlet CH4 and steam need to be preheated up to 350 ºC to carry out the SER stage al 650 ºC. The product gas is cooled to 35 ºC and subsequently compressed up to 30 bar to be purified in a PSA unit. A flow of 372 mol/s of pure H2 at 29 bar (stream 6) arises from the PSA, while 73 mol/s of an off-gas (stream 10) is obtained at near atmospheric pressure. The sensible heat of the O2-depleted air leaving the air reactor allows the mixture of carbonated particles and reforming catalyst coming from the SER stage (35.5 kg/s, stream 7) to be heated up to 711 ºC in a first preheating stage. Subsequently, the sensible heat of the CO2-rich gas obtained in the Fuel Reactor (containing 59.4 vol.% CO2) is sufficiently high to heat stream 8 up to 742 ºC while avoiding any carbonate decomposition (PCO2eq,720ºC=0.08). A minimum temperature of 870 ºC (PCO2eq,870ºC=0.65) in the Fuel Reactor is assumed in order to ensure the complete calcination of the calcium carbonate fed into this reaction stage (i. e., 6 ºC above the temperature corresponding to the equilibrium for 538 the CO2 partial pressure of 0.59 bar present at the fuel reactor outlet in the reference case). In this case, the heat flow required to carry out the operation is provided by a stream of 229 kg/s (mainly containing Fe2O3- based particles) arriving from the Air Reactor at 975 ºC. As can be seen in Figure 2b, lower S/C molar ratios reduce CH4 conversion to H2 and therefore lower H2 yields in the SER reactor are achieved. However, the reducing capacity of the PSA off-gas increases and a lower flow of additional CH4 is needed for the reduction of the Fe-based particles, improving the overall H2 yield of the process. Moreover, the use of the PSA off-gas as reducing agent highly improves the CO2 capture efficiency, since the carbon associated to the CH4 fed into the SER stage is finally emitted as CO2 easily purified that leaves the Fuel Reactor. For the reference case, a CCR of about 74 % would be achieved in case the PSA off-gas was not fuelled in the Fuel Reactor. Using this gas in the reduction/calcination stage the CCR increases up to 98 %. An adequate trade-off has to be found for the S/C molar ratio in SER stage in order to minimize the flow of additional CH4 fuelled in the Fuel Reactor while avoiding excessive costs associated to PSA operation. A more detailed heat integration of the process accompanied by an analysis of costs would be required, but this exercise is out of the scope of this work. At 870 ºC, the complete reduction of Fe2O3 to Fe3O4 is feasible. Moreover, H2 and CO fed into the Fuel Reactor will be totally oxidized to CO2 and steam, while CH4 can achieve a very high conversion of around 80 % (Adanez et al., 2012). As explained above, a solids flow rich in Fe3O4 (227 kg/s for the reference case, stream 15) is obtained in the segregator and subsequently directed to the Air Reactor. An excess of 7 vol.% of O2 is assumed for the oxidation stage to ensure the complete conversion of the Fe-particles to Fe2O3. Under these conditions, a flow of 364 mol/s of air needs to be supplied (stream 16), obtaining a flue gas of 293 mol/s containing 1.8 vol.% of O2 (stream 18). High proportions of inert in the iron-based particles will lead to lower oxidation temperatures in the Air Reactor, since inert phase will act as thermal ballast. However, there will be an increase in the heat transport capacity from the Air Reactor to the Fuel Reactor, at the expense of a higher circulation of solids between the reactors to fulfill the mass balance of the system. Consequently, the energy demand for the CaCO3 calcination will be reduced, improving the energy efficiency of the process (see Figure 3). For the reference case (i. e., 30 wt.% of Fe2O3 in the iron-based particles), a maximum temperature of 975 ºC is achieved. A solid/gas ratio of around 20 is needed in the Air Reactor in these conditions, which is within the usual operational range of large-scale circulating fluidized beds reported in the literature (Reh, 1995). Figure 3: Effect of the active content of the iron-based material on the energy efficiency of the process Table 2 summarises plant performance indexes obtained for the reference plants with and without CO2 capture (Martínez et al., 2014), together with those obtained for the proposed H2 production process combining SER with CLC. The FTR plant without CO2 capture operates with relatively low S/C molar ratios because the resulting off-gas from the PSA is burned in the FTR furnace to reduce the need of additional CH4 as fuel. In the reference FTR plant with CO2 capture, the S/C molar ratio must be significantly higher to promote the conversion of CH4 to H2 in the reformer and subsequently reach a high CO2 capture efficiency in the MDEA chemical absorption unit. In both plants, there is excess heat from the flue gases, which is recovered producing electricity through a steam turbine. For the commercial FTR plant, a ηeq,H2 of around 79 % is achieved, while the value for a FRT plant with MDEA absorption is significantly lower (around 70 %) due to the energy penalty associated to the CO2 capture system. In the proposed SER process combined with CLC, there is no excess heat available from the flue gases once steam and preheating gas requirements are fulfilled. However, a ηeq,H2 of 78.8 % (comparable to that obtained in a plant without CO2 capture) can be achieved while avoiding more than 98 % of carbon emissions to atmosphere, mainly due to the better H2 efficiency of the SER process and the very low energy penalties associated to the chemical looping system arranged for the regeneration of the CO2 sorbent. A more detailed integration of the process will be carried out in future works including a sensitivity analysis of the key design variables in order to determine the conditions that lead to optimal performance in terms of energy consumption and CO2 capture efficiency. 539 Table 2: Performance parameters associated to reference H2 production plants with and without CO2 capture and the proposed SER process combined with chemical looping combustion (after Martínez et al., 2014) Ref. FTR plant* FTR plant with CO2 capture* SER+CLC process S/C molar ratio 2.7 4.0 2.5 Thermal input, MW 121.9 130.8 - Steam turbine output, MW 3.3 3.8 - Air Separation Unit consumption, MW - - 0.1 CO2 compression, MW - 2.2 2.2 H2 compression, MW - - 5.7 Other plant auxiliaries, MW 1.0 1.4 0.4 Heat output, MW 8.6 4.1 - H2 yield in Reformer 3.1 3.6 3.5 Overall H2 yield 2.5 2.3 2.8 H2 production efficiency,% 73.9 68.8 85.2 Equivalent H2 production efficiency,% 79.0 70.0 78.8 Carbon capture ratio, % - 84.9 98.7 4. Conclusions A detailed process design of a novel H2 production process which combines SER and CLC has been described assuming a set of reasonable operating conditions. The presence of a CO2-sorbent during CH4 reforming enhances H2 production and simplifies its purification. A redox chemical loop for the calcination of the CO2-sorbent allows a separate stream of CO2 to be obtained with minimum energy penalty. H2 equivalent efficiencies up to 78.8 % (close to that obtained in commercial FRT plants) can be achieved with carbon capture ratios of around 98 %. These results demonstrate the potential of this process as a H 2 production technology with CO2 capture compared to conventional processes that uses commercial capture technologies. References Adanez J., Abad A., Garcia-Labiano F., Gayan P., de Diego L.F., 2012, Progress in chemical-looping combustion and reforming technologies, Prog. Energy Combust. Sci. 38, 215-282. Arias B., Diego, M.E. Abanades J.C., Lorenzo M., Diaz L., Martínez D., Alvarez J., Sánchez-Biezma A., 2013, Demonstration of steady state CO2 capture in a 1.7MWth calcium looping pilot. Int. J. Greenh. Gas Con. 18, 237-245. Barin I., 1989, Thermochemical Data of Pure Substances. VCH Verlagsgesellschaft, Weinheim, Germany. Fernández J.R., Abanades J.C., 2016, CO2 capture from the calcination of CaCO3 using iron oxide as heat carrier, J. Clean. Prod. 112, 1211-1217. Ge H., Shen L., Gu H., Song T., Jiang S., 2016, Combustion performance and sodium absorption of ZhunDong coal in a CLC process with hematite oxygen carrier, Appl. Therm. Eng. 94, 40-49. Gonzalez B., Grasa G., Alonso M., Abanades J.C., 2008, Modeling of the deactivation of CaO in a carbonate loop at high temperatures of calcination, Ind. Eng. Chem. Res. 47, 9256–9262. Harrison D.P., 2008, Sorption-enhanced hydrogen production: a review. Ind. Eng. Chem. Res. 47, 6486–6501. IEA, 2006, Hydrogen production and storage-R&D priorities and gaps, United Kingdom. IPCC, 2014, Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. Li G., Zhong G., Wu Q., 2016, Study on integrating a gas turbine in steam methane reforming process, Appl. Therm. Eng. 99, 919–927. Martínez I., Romano M.C., Fernández J.R., Chiesa P., Murillo R., Abanades J.C., 2014, Process design of a hydrogen production plant from natural gas with CO2 capture based on a novel Ca/Cu chemical loop, Appl. Energy 114,192–208. Naimer N.S., Chiba T., Nienow A.W., 1982, Parameter estimation for a solids mixing/ segregation model for gas fluidized beds, Chem. Eng. Sci. 37, 1047-1057. Reh L., 1995, New and efficient high-temperature processes with circulating fluid bed reactors, Chem. Eng. Technol. 18, 75-89. Romano M.C., Cassotti E.N., Chiesa P., Meyer J., Mastin J., 2011, Application of the sorption enhanced- steam reforming process in combined cycle-based power plants, Energy Proced. 4, 1125-1132. Yan L., Yue G., He B., 2015, Exergy analysis of a coal/biomass co-hydrogasification based chemical looping power generation system, Energy 93, 1778-1787. 540