Microsoft Word - PRES22_0064.docx DOI: 10.3303/CET2294068 Paper Received: 25 March 2022; Revised: 11 April 2022; Accepted: 12 April 2022 Please cite this article as: Monzer D., Bouallou C., 2022, CO₂ Capture and Purification for Use in the Molten Carbonate Electrolysis Cell, Chemical Engineering Transactions, 94, 409-414 DOI:10.3303/CET2294068 CHEMICAL ENGINEERING TRANSACTIONS VOL. 94, 2022 A publication of The Italian Association of Chemical Engineering Online at www.cetjournal.it Guest Editors: Petar S. Varbanov, Yee Van Fan, Jiří J. Klemeš, Sandro Nižetić Copyright © 2022, AIDIC Servizi S.r.l. ISBN 978-88-95608-93-8; ISSN 2283-9216 CO2 Capture and Purification for Use in the Molten Carbonate Electrolysis Cell Dayan Monzer*, Chakib Bouallou MINES ParisTech, PSL - Research University, CES - Centre d’Efficacité Energétique des Systèmes, 60 Boulevard Saint Michel, 75006 Paris, France Dayan.monzer@mines-paristech.fr The Molten Carbonate Fuel Cell (MCFC) is a technology that provides clean electricity and requires carbon dioxide to operate due to its carbonate electrolyte environment. This characteristic offers it an advantage as a technique for carbon capture during fuel cell operation and re-use of captured CO2 during electrolysis mode operation in the Molten Carbonate Electrolysis Cell (MCEC). In both cases, the cell’s feeds must be inspected for any impurities since MCFC/MCEC is highly sensitive to sulfur. Since CO2 as a feed for the cell is captured from the flue gases that contain impurities like SOx, NOx, and others, it is necessary to purify it to reach the MCFC’s tolerance limit. The purpose of this study is to simulate a full process for capturing CO2 and purifying it to fulfill impurity constraints of 1 ppm sulfur for use in MCEC. For this purpose, the conventional MEA-based absorption carbon capture method was created with a real flue gas composition input from a coal-fired power station and associated with two distinct desulfurization processes: wet limestone and cryogenic. According to this study, the wet limestone desulfurization process eliminates almost 100 % of the sulfur; however, the standard cryogenic process couldn’t attain a 1 ppm sulfur limit, promoting for studying a newly optimized cryogenic process with yielded 100 % of sulfur removal. As for the MEA-based absorption process, it captures around 95% of the CO2. The economic assessment revealed that the desulfurization process based on the optimized cryogenic process combined with the MEA capture system provides a lower total cost than the one based on the wet limestone. The latter's high expenses contribute to the requirement of an excessive amount of raw material to attain the threshold of 1 ppm sulfur. This overall analysis demonstrates that the cryogenic approach appears to be the best option for this target, whereas the limestone process requires more improvements to reduce its costs. 1. Introduction In the context of increasing attention to global warming, several countries, such as Germany, signed the Paris Agreement on April 22, 2016. The main goal of this agreement is to reach carbon neutrality in the second half of the century, to limit the environmental and social consequences of an excessive increase in the earth's temperature. To achieve this goal, the German government has put in place the Climate Action Plan 2050. Among the several objectives set by the German government to achieve carbon neutrality by 2050, the industrial sector must reduce its emissions from 49 % to 41 % by 2030, compared to 1990. To achieve this, one of the measures put forward by the German government is the establishment of a research and development program to reduce greenhouse gas emissions from traditional industrial processes. It is in this context, that the capture of CO2 from the flue gases of certain fossil fuels seems necessary. Given that Germany is still a major user of coal, ranking seventh in the list of countries with the most coal-fired power plants in operation in 2021, it seems interesting to consider a method for capturing CO2 from such power plants. However, this issue of CO2 cannot be addressed without thinking of its future use. The most relevant way to valorize the CO2 captured from flue gases would be the use of MCFCs, a cell that requires CO2 with its feed for operation to produce and store clean electricity. Indeed, by using CO2 as a reagent, this cell allows the production of electricity when used in fuel cell mode, and produces, when used in electrolysis mode (MCEC), synthesis gases which allow the storage of electricity produced by renewable energy. The detailed operation of the MCEC is fully described in prior publication by Monzer et al. (2021). 409 However, the CO2 captured from flue gases that will be used as a feed for the cell contains impurities including SOx, NOx and others to which the MCFC has been reported to be highly sensitive. This implies the necessity to purify the feed stream in order for it to reach the MCFC’s tolerance limit which remains quite unclear for now. Indeed, although sulfur has been recognized as the most harmful impurity in the flue gas for the MCFC (Di Giulio et al., 2012), operating conditions such as temperature or pressure seem likely to modify the influence of sulfur on the cell degradation and the tolerance limit of the cell. A broader look at all the impurities that are likely to be found in flue gas is needed to avoid cell deterioration. Given that the purpose of the work reported in this research paper study is to simulate a full process for capturing CO2 and purifying it to fulfill impurity constraints for use in the MCEC cell, it is necessary to find an appropriate desulfurization mode for this situation. A thorough literature review carried by Li et al. (2022) led to the selection of two desulphurisation methods, which had already been implemented on the ASPEN Plus software: cryogenic desulfurization and desulfurization using limestone. However, none of these simulations were focused on reducing the sulfur concentration to 1 ppm, nor were they combined with a CO2 capture system. As a result, for the method of capturing CO2 from the flue gas, it was decided to investigate a chemical absorption method using MEA as a solvent. This process; however, is sensitive to sulfur with a tolerance limit of 10 ppm. So, it seemed appropriate to perform the desulfurization to 1 ppm before capturing CO2 for its final use in the MCEC cell. Several models of the MEA capture process have already been set up using ASPEN Plus (Alie, 2004), but in each of these models, it was assumed that the incoming gas no longer contained any sulfur. 2. Methods 2.1 Flue gas composition To begin, as the main goal of the research is to simulate the whole purification and capture process, as close to reality as possible, the real composition of the feeding flue gas needs to be found. However, since the biggest impacts are related to CO2 and impurity concentration, finding a complete composition that did not compromise the precision of the concentration of impurities was impossible. This led to the selection of the volumetric percentages of most components such as CO2 and all impurities from Lee et al. (2009) experiments, taking classic concentrations of O2 and H2O from various sources in literature such as Song et al. (2004) and Di Giulio et al. (2012) and using N2 as balance, as it is an inert gas. Table 1: Flue gas composition and Simulation Parameters Parameters Value Chemical species Volumetric percentage Inlet Flue Gas Parameters CO2 12 Temperature (°C) 52 SO2 0.3 Pressure (MPa) 0.1 SO3 0.006 Volumetric flow rate (m3/h) 2.55x106 NO2 0.004 Process Parameters HCl 0.01 MEA inlet mass fraction 0.3 H2O 10 Lean MEA purge loss (%) 15 O2 5 CaCO3 inlet mass fraction 0.0007 N2 72.68 2.2 Tolerance limit for impurities In relation to the composition that has been chosen to simulate a flue gas from a coal-fired power plant, it seems important to look at the impact that different impurities have on the cell, in order to set a limit to be respected thanks to a prior purification of the gases. Since MCEC and MCFC cells have the same chemical composition but simply a different electrochemical functioning, it will be adequate to look at the impact of the distinct species on one of these two cells: the MCFC. Di Giulio et al. (2012) took up the existing body of work, recognized that sulfur is the most dangerous impurity in the flue gas for the MCFC and studied the impact of operating conditions on the tolerance limit of this cell to sulfur. Sulfur interacts with the cell in the following way: SOx(g) contained in flue gas is readily uptake by the molten carbonate electrolyte. The molecules Na2SO4 and Li2SO4 formed are highly soluble in the molten carbonate, and migrate within the electrolyte. At the anode, these species are reduced and eliminated from the electrolyte in the form of H2S(g), where it may interact with the anode. The results of their work showed that the temperature has a major impact on the degradation of the cell due to sulfur and that such degradation can occur from 1 ppm H2S. Subsequently, Milewski and Lewandowski (2012) investigated the impact of sulfur in high concentrations up to 680 ppm, closer to the concentrations found in the flue gases. They concluded that even for concentrations lower than 100 ppm, a voltage drop across the cell of 2 % is observed after 48 h. It can be deduced that such concentrations would seriously damage the cell very quickly, severely reducing its operational life. Marina (2014) takes a broader look at all types of impurities, 410 including sulfur, that are likely to be found in combustion flue gases. This paper states that negligible performance losses are observed for sulfur concentrations up to 1 ppm. By studying the other impurities such as mercury, chlorine, fluorine and selenium (Marina, 2014), it can be concluded that only chlorine could poison the cell. Chlorides are favored to displace carbonate from the molten salt electrolyte. Chlorides are then eliminated from the electrolyte through volatilization of alkali chlorides, which could lead to substantial electrolyte loss. Chlorine can be found at concentrations up to 0.1 ppm HCl(g): from this limit, 5 % electrolyte loss would be observed during the cell's operating time. It will only be necessary, given the above composition, to desulfurize the smoke from the thermal power plant. 2.3 Desulfurization process Li et al. (2022) in their paper, provided a state of the art of desulfurization. He highlighted the fact that among dry, wet, and semi-dry desulfurization, wet desulfurization is the most efficient. In addition, Gutiérrez Ortiz et al. (2006) argue that desulfurization using limestone is the most suitable for the current study’s purpose due to its low installation cost. Figure 1 shows a scheme of the wet limestone desulfurization process. The flue gas desulfurization happens in the absorber tower where the slurry – a calcium carbonate mixture – reacts with the SOx and removes it by chemical absorption. A filter is added after the absorber tower to remove the precipitated gypsum and recycle the liquid outlet through a loop. In order to reach the goal of 1 ppm sulfur at the outlet, an optimal calcium carbonate to sulfur ratio needed to be found. A sensitivity analysis on ASPEN Plus demonstrated that the more slurry entering, the more efficient the desulfurization. However, the Economic Analysis module revealed that the most costly component of this process is the raw material (CaCO3) contained in this slurry leading to the selection of the lowest mass flow, allowing it to reach the objective of 1 ppm. Figure 1: ASPEN Plus implementation of limestone desulfurization In addition, recent research led by Zhang et al. (2021) has focused on the use of a new process: cryogenic desulfurization. Their simulation results demonstrate that this process, which is based on the liquefaction of sulfur in the flue gas, provides a desulfurization efficiency of 73.2 % at a low temperature of -45 °C and pressure of 0.5 MPa where this efficiency can reach 90.2 % with a pressure increase to 2.5 MPa. Based on this study, the cryogenic process was adapted and modified to desulfurize the flue gas to a target of 1 ppm. However, CO2 liquefies with sulfur at high pressures based on phase diagram data which were the condition cases addressed by Zhang et al. (2021). As a result, the process's working pressure has been reduced to 0.1 - 0.2 MPa to avoid CO2 liquefaction with sulfur. The standard cryogenic desulfurization process was studied at low pressure as well as an optimized process was proposed and studied. The latter consists of an additional pre-reactor that was added to shift SO2 into SO3 since SO3 has a higher boiling point than that of SO2, as illustrated in Figure 2. Figure 2: ASPEN Plus implementation of cryogenic desulfurization The general operation of the current process is to primarily cool down the gas stream to remove the excess water stream as the subsequent temperature will be very low and would freeze the water, which would probably 411 block the pipes. At this stage, two successive heat exchangers were introduced to lower the temperature to 10 °C and separate the water from the rest of the gas stream. The outlet vapor stream of the separator (SEP1) is subsequently cooled to -90 °C, where the sulfur is separated from the rest of the flue gas. The latter vapor stream from the separator (SEP2) is reheated by the inlet feed using the second stage heat exchanger, as the temperatures required for the subsequent CO2 capture are higher which is selected to be 25 °C in this work. 2.4 CO2 capture process CO2 capture can take several forms, from pre-combustion to post-combustion capture. The advantage of post- combustion capture is that it would be compatible with existing thermal power plants, being built on top of them. There has been a lot of research into the chemical capture of CO2 by chemical absorption using different solvents. The most mature solvent for this process is monoethanolamine (MEA), but other solvents were also studied and revealed an interesting performance. For example, Molina and Bouallou (2015) have studied the CO2 absorption using ammonia solvent that has proved the possibility of capturing CO2 with an efficiency of 90 %. Regardless of the solvents investigated, the focus of this work has been on simulating the reference MEA CO2 capture process. Such a process has already been modeled on the ASPEN Plus software (De Castro et al., 2010), with the aim of achieving a CO2 rate of 85 % in the outgoing flows. However, these studies did not take into account the presence of sulfur traces in the gas. In this study, two processes were considered, a standard MEA capture process (Figure 3a) and another process considering the heat recovery from the process (Figure 3b) where two heat exchangers (HeatX1 and HeatX2) were added in the recycling loop to replace the cooler in Figure 3a. Figure 3a: Carbon capture without Heat Exchangers Figure 3b: Carbon capture with Heat Exchangers 3. Results 3.1 Flue Gas Results Five different simulations were simulated using ASPEN Plus software: the flue gas desulfurization using the wet limestone process at 0.11 MPa, the cryogenic desulfurization in two different processes at 0.2 MPa, and two different CO2 captures of the flow from the optimized cryogenic desulfurization (with reactor). The flue gas composition at each stage of the two desulfurization processes are shown in Table 2. Table 2: Flue gas composition at various stages of the wet limestone and the cryogenic desulfurization processes Chemical species Entering Flue Gas Mass Fraction Outgoing Flue Gas Mass Fraction (wet limestone) Outgoing Flue Gas Mass Fraction (cryogenic) Without reactor With reactor CO2 0.1610 0.164 0.1609 0.1644 SO2 0.0067 1.32x10-24 0.0023 1.35x10-13 SO3 0.00019 0.269x10-6 0.0126x10-6 0.77x10-6 NO2 9.385x10-5 3.238x10-7 1.52x10-6 2.03x10-6 HCl 0.00011 0.000126 0.000135 0.000135 H2O 0.163 0.077 9.99x10-18 1.3x10-17 O2 0.05 0.053 0.061 0.06 N2 0.619 0.704 0.776 0.776 The limestone desulfurization process demonstrated that it could fulfill the sulfur tolerance threshold of 1 ppm. However, because of the limitations of CO2 liquefaction, the standard cryogenic technique (without reactor) 412 couldn’t achieve a sulfur level of 1 ppm where the SO2 outlet mass fraction is 230 ppm. A pressure increase to 3 MPa improved the desulfurization process to 1 ppm sulfur. This condition, however, results in a massive loss of CO2 with sulfur in the form of liquid. As a result, the standard cryogenic technique appears ineffective for desulfurization to 1 ppm sulfur. On the contrary, the addition of the pre-reactor has improved the cryogenic desulfurization where it can be found in Table 2 that the sulfur mass fraction in the outgoing flue gas is below 1 ppm. In a comparison of both desulfurization processes, the wet limestone process achieved almost 100 % yield of both SO2 and SO3 removal, whereas the standard cryogenic one (without reactor) yielded 72 % and almost 100 % of SO2 and SO3 removal. Nevertheless, the optimized cryogenic process (with reactor) proves almost 100 % yield of both SO2 and SO3 removal which could be competitive with the conventional wet limestone process. However, a 31 % CO2 loss was obtained in the optimized cryogenic process which requires further studies to reduce or avoid it. For MEA, both simulations, whose flue gas composition can be found in Table 2, also reach their goals which are on one hand to capture a large part of the CO2 from the flue gas to release carbon-free fumes into the atmosphere, and on the other hand, concentrate the CO2 to have an outlet flow close to purity to use as feed for the cell. The post-limestone MEA process yielded a 94.5 % and 94.7 % CO2 capture for the process without and with recovery heat exchangers (HeatX), respectively. On the other hand, the post-cryogenic MEA process achieved a 94.8 % CO2 capture yield with both processes. Besides, the reboiler duty proved a decrease of 20 % with the recovery heat exchangers (HeatX) optimized process for the post-limestone MEA process and a decrease of 36 % for the post-cryogenic MEA process. 3.2 Cost Calculations Another element to consider is the cost of these different processes which permits to compare them in terms of economic aspects. The cost is divided into two main categories. On one side, there is the initial investment (CAPEX) which includes the equipment and installation costs and on the other side there is the operating costs (OPEX) which include all ongoing costs such as the raw materials used and the electricity. The Economic Analyzer tool on ASPEN Plus calculates the equipment and installation costs of most blocks used in the different processes. However, it does not take all of them into account and needs to be completed using the data in Jones (2019). The operating costs can be of numerous forms, such as energy costs, raw material costs, and labor costs. The Energy Analyzer program estimates total energy expenses, while labor and raw material costs are determined based on market prices. In the current case, OPEX represents the sum of total utilities (energy expenses) and raw material costs. The raw material cost is calculated using Eq(1) based on the requirements of the raw materials of each process, where t is the operating time in hours (h), M is the molar mass in kg/kmol, F(i) is the mole flow of the raw material stream(i) in kmol/h, and C is the cost in USD/kg. 𝑅𝑎𝑤 𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙 𝑐𝑜𝑠𝑡 = ∑ 𝐹(𝑖) × 𝑥(𝑖) × ∆𝑡 × 𝑀(𝑖) × 𝐶(USD/𝑘𝑔𝑖 ) (1) Figure 4: Total actualized cost of both desulfurization processes, along with the MEA-absorption process 3.3 Discussion The yield calculations revealed that the desulfurization efficiency is analogous in the wet limestone and the modified cryogenic process. To compare the two techniques, the most relevant differentiating factors would be the total costs, so they were investigated, as shown in Figure 4. First, there is the cost specific to desulfurization. On the CAPEX side, the wet limestone desulfurization process investment is lower by 85 % than the optimized cryogenic process, where the investment cost of the cryogenic process is concentrated around the heat exchangers and the reactor. On the other side, both techniques have a significant operating expenses, with the wet limestone process being very demanding in raw materials in the entering slurry and the cryogenic process being very energy-consuming. Nevertheless, the optimized cryogenic process requires a lower operating cost 413 of 43 % than that of the wet limestone process due to the availability of heat integration in the cryogenic process that lowers the energy duty, but there is no way to reduce the raw material demand in the limestone process at the moment. Secondly, there is the cost related to CO2 capture. The CAPEX in the capture post-limestone may be somewhat higher by an average of 29 % than in the capture post-cryogenic; however, the operational expenses in the capture post-limestone are far higher by an average of 40 %. This can be explained by two main factors. First, the entering flue gas needs to be compressed and cooled, which is not the case of post- cryogenic, and that adds an energy operating cost. Secondly, the flow of the MEA entering the absorption tower needed to reach the tolerance limit is much bigger, which adds a raw material cost. When considering the two different processes to capture CO2 (with and without HeatX), it can be seen that while the process with heat exchangers has a higher installation cost, there is no significant improvement in the OPEX in the post-limestone case due to the large energy consumption for compression and cooling. However, the post-cryogenic proved a 4 % decrease in the OPEX when adding heat exchangers. Based on this economic study, the most viable option is to choose cryogenic desulfurization as the operating costs for the wet limestone are far too great. However, as the operating costs for the limestone process are mostly related to raw materials, more research on how to recycle the slurry and the MEA could improve the OPEX. 4. Conclusion In this work, two different flue gas desulfurization and CO2 capture processes were simulated using ASPEN Plus to obtain an acceptable CO2 gas for the MCEC cell. Based on this research, it can be concluded that under the right operating conditions, both the wet limestone and optimized cryogenic desulfurization processes can produce flue gases containing less than 1 ppm of sulfur, whereas the standard cryogenic process was unable to achieve the 1 ppm target limit. While the wet limestone process is interesting relative to its maturity, the proposed cryogenic process provides comparable results with lower operating costs requirements. This cryogenic process combined with the MEA CO2 capture process also led to a lower overall cost (CAPEX+OPEX) than the wet limestone process coupled with the MEA CO2 capture. This assessment reveals that the cryogenic process seems the most suitable option coupled with MEA to have a CO2 with less than 1 ppm sulfur. References Alie C.F., 2004, CO2 Capture With MEA: Integrating the Absorption Process and Steam Cycle of an Existing Coal-Fired Power Plant, MSc at University of Waterloo, Ontario, Canada. De Castro J., Rivera-Tinoco R., Bouallou C., 2010, Hydrogen Production from Natural Gas: Auto-Thermal Reforming and CO2 Capture, Chemical Engineering Transactions, 21, 163-168. Di Guilio N., Bosio B., Cigolotti V., Nam S. W., 2012, Experimental and Theoratical Analysis of H2S Effects on MCFCs, International Journal of Hydrogen, Energy, 37, 19329-19336. Gutiérrez Ortiz F. J., Vidal F., Ollero P., Salvadro L., Cortés V., 2006, Pilot-Plant Technical Assessment, of Wet Flue Gas Desulfurization Using Limestone, Industrial and Engineering Chemical Research, 45, 1466-1477. Jones D. A., 2019, Technoeconomic Evaluation of MEA versus Mixed Amines and a Catalyst System for CO2 Removal at Near-Commercial Scale at Duke Energy Gibson 3 Plant and Duke Energy Buck NGCC Plant, Technical Report, LLNL-TR-767484. Lee J. Y., Keener T. C., Yang Y. J., 2009, Potential Flue Gas Impurities in Carbon Dioxide Streams Separated from Coal-fired Power Plants, Journal of the Air and Waste Management Association, 59, 725-732. Li X., Han J., Liu Y., Dou Z., Zhang T, 2022, Summary of Research Progress on Industrial Flue Gas Desulfurization Technology, Separation and Purification Technology, 281, 119-849. Marina O. A., 2014, Molten Carbonate Fuel Cell Interactions with Impurities in Synthetic Flue Gas, Technical Report, PNNL-28043. Milewski J., Lewandowski J., 2012, Separating CO2 from Flue Gases Using a Molten Carbonate Flue Gas, IERI Procedia, 1, 232-237. Molina C. T., Bouallou C., 2015, Carbon Dioxide Absorption by Ammonia Intensified with Membrane Contractors, Chemical Engineering Transactions, 45, 253-258. Monzer D., Rivera-Tinoco R., Bouallou C., 2021 Investigation of the Techno-Economical Feasibility of the Power-to-Methane Process Based on Molten Carbonate Electrolyzer, Frontiers Energy Research, 9, Article 650303. Song C., Pan W., Srimat S. T., Zheng J., Li Y., Wang Y. H., Xu B. Q., Zhu Q. M., 2004, Tri-reforming of Methane over Ni Catalysts for CO2 Conversion to Syngas with Desired H2CO Ratios Using Flue Gas of Power Plants without CO2 Separation, Studies in Surface Science and Catalysis, 153, 315-322. Zhang Y., Wang Y., Liu Y., Gao H., Shi Y., Lu M., Yang H., 2021, Experimentals and Simulation of Varying Paramaters in Cryogenic Flue Gas Desulfurization Process on ASPEN Plus, Separation and Purification Technology, 259, 118-223. 414