CET Volume 86 DOI: 10.3303/CET2186168 Paper Received: 16 August 2020; Revised: 12 March 2021; Accepted: 6 April 2021 Please cite this article as: Tregambi C., Bareschino P., Mancusi E., Pepe F., 2021, Development and Techno-Economic Analysis of a Two Carriers Reactor Arrangement for Chemical-Looping Combustion in a Fixed Bed, Chemical Engineering Transactions, 86, 1003-1008 DOI:10.3303/CET2186168 CHEMICAL ENGINEERING TRANSACTIONS VOL. 86, 2021 A publication of The Italian Association of Chemical Engineering Online at www.cetjournal.it Guest Editors: Sauro Pierucci, Jiří Jaromír Klemeš Copyright © 2021, AIDIC Servizi S.r.l. ISBN 978-88-95608-84-6; ISSN 2283-9216 Development and Techno-Economic Analysis of a Two Carriers Reactor Arrangement for Chemical-Looping Combustion in a Fixed Bed Claudio Tregambi, Piero Bareschino, Erasmo Mancusi*, Francesco Pepe Dipartimento di Ingegneria, Università degli Studi del Sannio, Piazza Roma 21, 82100 Benevento, Italia erasmo.mancusi@unisannio,it Reduction of CO2 emissions is imperative to stop climate change. In the present study, the performance of a system based on chemical looping combustion in fixed bed is investigated. The plant allows power generation with obtainment of a CO2-rich stream ready for sequestration. To overcome the problems related to large temperature variations, use of two in-series oxygen carriers, Cu/CuO and Ni/NiO supported on Al2O3, is investigated. CH4 was considered as fuel. A 1D numerical model was developed to estimate temperature and concentration profiles within the fixed bed as a function of time. A network of parallel reactors was designed to obtain continuous power generation at the turbine. A techno-economic analysis was performed to estimate plant throughput, overall efficiency, total capital costs, and levelized cost of energy of the proposed system. 1. Introduction Carbon dioxide has been recognized as one of the main contributors to global warming. Carbon capture and sequestration technologies such as oxyfuel combustion, calcium looping and chemical looping combustion (CLC) may represent viable options to process conventional and/or renewable fuels in a clean way, allowing for production of concentrated CO2 streams ready for subsequent sequestration (Adànez et al., 2012, Tregambi et al. 2020) or reutilization (Tregambi et al. 2021). CLC splits the conventional combustion in a two steps process. In the first stage, the fuel is reacted with a solid material which provides the O2 required for the fuel oxidation. Since air is not fed, exhaust gas is not diluted with N2 and a stream of pure CO2 is produced upon water condensation. In the second step, the solid material is oxidized back with an air stream. Exhaust gas consists of O2-lean air and can be released to the atmosphere upon heat recovery. The solid carrier usually consists of a metal oxide, referred as oxygen carrier (OC). First step of the process (fuel oxidation with OC reduction) can be endothermal or exothermal according to the fuel/OC couple, whereas the oxidation of the OC is always exothermal (Abad et al., 2007). Most of existing CLC reactors are based on two interconnected fluidized beds, one of them acting as the fuel-reactor and the other one as the air-reactor (Diglio et al., 2017a). Alternatively, a packed bed reactor technology for CLC has been proposed (Noorman et al., 2007). In a packed bed, the OC particles are stationary and are alternately exposed to reducing and oxidizing conditions through a periodic switching of the feed conditions. The choice of the type of reactor and of the OC is crucial. Indeed, to achieve a high electricity efficiency the flue gas to be expanded in turbine needs to be produced at 20-30 bar and 1200 °C (Hamers et al., 2015). Fixed beds are more easily operated up to the high-pressure values required for subsequent gas turbine expansion of exhaust gas, and also reduce the problems of attrition/elutriation of the solid reactive material (Noorman et al., 2007). To circumvent the problem of maximum temperature achievable, Hamers et al. (2015) and Kooiman et al. (2015) proposed a two stage-CLC (TS-CLC) using the pair Cu/Mn in the first case and Cu/Ni in the second one. In both studies, syngas was used as fuel. An alternative option is represented by use of methane as fuel (Diglio et al., 2018). The aim of the present study is to investigate a CLC process in fixed bed reactors by exploiting the peculiarities of two-stage CLC, starting from material/energy balance toward techno-economic analysis. Methane was considered as fuel, whereas Cu/CuO and Ni/NiO were selected as active phase for the OCs. First, a transient model featuring heat and mass balance equations was developed to investigate the key 1003 performance of the system. Then, given the intrinsic intermittent nature of fixed beds, an integrated reactor network was designed with the aim of ensuring continuous power generation at the turbine. Finally, a techno economic analysis was performed to assess capital costs and levelized cost of energy. 2. Methodology 2.1 Mathematical model of chemical looping combustion CLC process consists in a pressurized fixed bed with two in series OCs, a Cu-based OC in the first section, and a Ni-based OC in the second section. The reaction kinetic model for Cu/CuO (Abad et al., 2007) and Ni/NiO (Iliuta et al., 2010) carriers on γ-Al2O3 is summarized in Table 1. Table 1: Kinetic scheme adopted Reaction ∆H0, kJ·mol -1 Reaction ∆H0, kJ·mol -1 2Cu + O2 → 2CuO -296 R1 NiO + CO ⇌ Ni + CO2 -43 R6 2Ni + O2 → 2NiO -479 R2 NiO + CH4 ⇌ Ni + CO + 2H2 203 R7 4CuO + CH4 → 4Cu + CO2 + 2H2O -209 R3 CH4 + H2O ⇌ CO + 3H2 206 R8 2NiO + CH4 ⇌ 2Ni + CO2 + 2H2 161 R4 CH4 + CO2 ⇌ 2CO + 2H2 247 R9 NiO + H2 ⇌ Ni + H2O -2 R5 CO + H2O ⇌ CO2 + H2 -41 R10 During oxidation stage (OS), air is fed and R1 and R2 occur in the first and second section of the reactor, respectively. Then, CH4 is used to reduce CuO and NiO. During reduction stage (RS), the exothermic CuO reduction to Cu (R3) occurs in the first section, while in the second section endothermic NiO reduction (R4-R7) occurs simultaneously to CH4 reforming (R8-R9). A detailed kinetic expression of R1 and R3 can be found in Abad et al. (2007), while the Ni oxidation (R2) and reduction reaction rates are well presented by Iliuta et al. (2010). To describe axial concentration and temperature profiles in the reactor, a 1D pseudo-homogenous packed-bed model was used. The absence of radial concentration and temperature gradients, as well as the lack of both interphase and intra-particle concentration and temperature gradients has been validated and more details can be found in Mancusi et al. (2020). Governing equations for both OCs are reported in Table 2. Table 2: Governing equations Description Equation Gas phase mass balance + = + Solid phase mass balance = Energy balance + 1 − + = + −Δ Momentum balance − = 150 + 1.75 Boundary conditions ( , ) = (0, ) − , , ( , ) = 0 ( , ) = ( (0, ) − ) , ( , ) = 0 In Table 2, the i index represents the gaseous species (i=CH4, H2, CO2, H2O, CO, O2, N2), while k (k=Ni, Cu) the solid carriers. The rates of formation or consumption of gas (rj) and solid (rk) species were calculated by summing up the reaction rates of those species in the reaction Ri in Table 1 for i=1,…,10. C is the gas concentration in mol·m-3, C0k is the initial concentration of solid species in the carrier, T is the temperature in K and X is the solid conversion, P is the pressure in Pa, z is the axial variable in m, t is the time in s, εg is the bed void fraction, ug is the gas superficial velocity in m·s -1, Dax is the axial dispersion in m 2·s-1, ρoc and ρg are the density of oxygen carrier and gas, respectively, in kg·m-3, cpg is the gas heat capacity in J·kg -1·K-1, λeff is the effective thermal conductivity in W·m-1·K-1, dp is the particle diameter in m, ΔH is the reaction enthalpy in kJ·mol-1. TS-CLC was modeled as two in-series fixed beds. Exit conditions of temperature and concentration from the first reactor represent the inlet to the second one. The infinite dimensional partial differential equations (PDEs) system has been reduced to a set of 400 ordinary differential equations (ODEs) by finite difference techniques which have been numerically solved by making use of the fortran library DLSODES (e.g. Altimari et al., 2012). 1004 2.2 Design of the integrated system and techno economic analysis Operation of the TS-CLC involves a sequence of five steps: i) oxidation stage (OS), where air is fed to the reactor for the OC oxidation; ii) heat removal (HR) step, during which air is fed to the reactor to extract the heat trapped by the OC as a consequence of the OS, and the resulting high-temperature gas stream is sent to turbine for power generation; iii) reduction step (RS), during which the fuel is fed to the reactor to reduce the OC, completing the looping process. Two purge steps (PS) are required between stages ii) and iii) and after stage iii) to avoid formation of potentially explosive mixtures. In conclusion, the TS-CLC includes RS-PS-OS- HR-PS that cyclically follows each other in the fixed bed. The stream of N2 produced after OS can be recycled in the HR step. The stream of CO2 and H2O after RS is sent to storage upon water condensation. To ensure continuous power generation at the turbine, a configuration of multiple in-parallel fixed beds needs to be designed, for which different approaches can be followed. Hamers et al. (2015) coupled CLC in fixed beds with an integrated gasification combined plant, meaning that a continuous stream of syngas needed to be processed. In this work, the fuel is CH4 and it is considered to be available on demand. The reactor network was designed with the aim of obtaining a constant power production at the turbine, and with the two further goals of: i) keeping the investment costs as low as possible, and ii) avoiding the use of a gas buffer prior to the turbine. For this to occur, at least one reactor should always be within HR step, and it should be verified that: ∙ = 1 (1) Where Nr the number of in-parallel reactors, τHR is the duration of the HR step and τTOT that of a whole cycle. Net power production (Pnet) was estimated as difference between the power produced at the turbine and the sum of that required by the compressors and that possibly needed to preheat the reactants. Compression and expansion were modelled as single stage isentropic processes. An overall efficiency factor ( ) was defined as: = (2) where WRS is the stream of CH4 fed during RS, the lower heating value of CH4 and τRS the RS duration. The rector network was designed by accounting for the main plant components, namely: air/methane compressors, fixed bed reactors with OCs, high-temperature outlet valves for gas switching, gas turbine. A sketch of the system is presented along with discussion of results, as a complete design is possible only upon solution of the mathematical model of CLC. Total capital costs (TCC) were then evaluated as sum of the individual costs of the different components. Costs of reactors and high temperature valves were computed according to Hamers et al. (2015), considering that fixed bed reactors embody an internal and external refractory, and a steel vessel. Cost of the individual materials was evaluated, and their sum multiplied by 4 to account for the reactor effective construction. OCs cost was estimated as sum of the individual metal oxides and inert support costs (Cu, Ni, γ-Al2O3), and again multiplied by 4 to account for the synthesis procedure. For turbine and compressors, equations from literature were used. Finally, levelized cost of energy was estimated as: = ∙ ∙ ∙ + + (3) where FCF is the fixed charge factor, computed considering 25 years of operation and a project interest ratio of 8.75%. Fixed operating and maintenance costs (FOM) were assumed to be 1% of the TCC and the capacity factor (CF) was set equal to 0.85. Variable operating and maintenance costs (VOM) were evaluated by considering replacement of OCs and of high temperature valves, as well as the cost related to transport and storage of CO2. Specific cost of the fuel (SFC) was considered. 3. Results and discussion The core concept of TS-CLC is to transfer the heat developed during each stage in the first OC to the second one, so that the desired temperature increase can be gradually attained. The results of the numerical simulation of the TS-CLC are discussed below, and the parameters used are reported in Table 3. Table 3: Model parameters used for TS-CLC numerical simulation Parameter Value Parameter Value Parameter Value Pin, Pa 20∙106 LCu, m 1.0 wactNi0 0.11 Tin,OS, °C 450 LNi, m 1.0 wactCu 0 0.14 Tin,RS, °C 550 dr, m 0.7 mCu, kg 78.4 T0, °C [450,950] εg 0.6 mNi, kg 58 | / , kg·m-2·s-1 2 | , kg·m-2·s-1 2/15 | , kg·m-2·s-1 4 1005 In Figure 1 are shown several spatial profiles during the main stages (RS, OS and HR) at the beginning and at the end of each stage. Since the Cu reduction reaction (R2) is weakly exothermic while the Ni reduction stage is endothermic (R4-R9), in the first part of reactor the temperature increases, decreasing in the second one (see Figure 1a). The heat produced during RS in first reactor (Cu-based carrier) is then transferred to the second reactor, containing Ni-based carrier, to mitigate the temperature decrease due to the previous RS. Cu and Ni oxidation are both exothermic (R1-R2) and a sharp increase in the temperature is observed upon OS (Figure 1b). Once OS is completed the HR occurs and the heat produced during the previous OS is swept away and sent to the turbine. To power the turbine with an almost constant temperature HR is extended until outlet gas temperature drops below 1160 °C. More details about the control strategy that dictates the switch between each stage can be found in Mancusi et al. (2020) and Diglio et al. (2017b). Figure 1: Spatial temperature profiles at several time instants during RS (a), OS (b) and HR (c) Finally, in Figure 2 the outlet gas temperatures at the outlet of first and second reactor (a) and the O2 and N2 molar fractions (b) are reported for several CLC cycles when the regime conditions are attained. Figure 2: Outlet gas temperature (a) and N2-O2 concentrations (b) vs time for first and second OC. It is possible to see that the temperature increase of 600 °C between inlet and outlet is equally split between the two oxygen carriers. The length of each stage was not fixed a priori, but a controller automatically sets it. The time lengths found using the parameters set previously reported (see Table 3) are detailed in Table 4. Table 4: Period of each stage Stage Oxidation Purge Reduction Heat Removal Purge Overall Value (s) 210 30 205 475 30 950 In order to produce a continuous hot gas stream to power the turbine, a system featuring in parallel reactors has to be operated. From Eq. (1) it is computed that Nr equals 2 for the investigated TS-CLC, as duration of the HR step equals half the overall cycle length. Therefore, an overall integrated scheme featuring two in parallel reactors was designed to attain a continuous power generation. Figure 3 depicts the process scheme 1006 designed for the CLC operation. Only one of the reactors is sketched, the other working exactly in the same way but delayed in time. Thermal buffers, turbine, and compressors are shared between the two reactors. Figure 3: Sketch of the reactor network for the TS-CLC. HTV = high temperature valve, TB = thermal buffer Preheating of the reactant streams can be performed by exploiting the sensible heat of the products streams, i.e. no auxiliary heat is required for the operation of the system and the whole power produced by the turbine can be sold to the market. More into detail, preheating of the air stream required for the OS and HR can be ensured by exploiting the sensible heat of the nitrogen stream exiting the OS. After heat recovery, the pressurized N2 stream is recycled in the HR step to reduce power consumption of the compressor during the HR step. The same heat recovery strategy applies for preheating of the methane stream: sensible heat of the CO2+H2O stream produced after the reduction step can be recovered in a thermal buffer for the purpose. Finally, also for the PS the sensible heat of the impure N2 exiting the system is used to preheat the stream of fresh N2. Analysis of the power produced by the turbine and required by the compressors, not detailed here for the sake of brevity, disclose that the integrated reactor network can produce about 241 kWe of energy, with an overall efficiency of 22%. Capital costs of the proposed integrated plant are instead detailed in Table 5. Table 5: Total capital costs of the integrated plant Component High temperature valves Reactors + oxygen carriers Compressors Turbine Total Cost (k€) 174.3 20.7 115.2 113.4 423.7 Total capital costs equal about 424 k€. The most expensive component are the high temperature valves required at the outlet of the fixed bed reactor, which account for about 41% of the TCC. Compressors and turbine account each for 27% of the total capital costs, whereas the fixed bed reactors and the oxygen carrier represent the smallest fraction of TCC, accounting for merely 5%. Finally, levelized cost of energy is reported in Figure 4, split in the three main contributions related to: i) fuel; ii) variable operating and maintenance costs; iii) capital costs and fixed operating and maintenance costs. Figure 4: Levelized cost of energy for the TS-CLC, split in the three main contributions The overall levelized cost of energy values about 134 €/MWhe. This value is larger than that of conventional power plant or integrated gasification combined cycle based on coal or methane without carbon capture and 93.2 27.6 22.1 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 Fuel TCC + FOM VOM Levelized cost of enegy [€/MWhe] 1007 storage, which generally ranges within 30–60 US$/kWhe. However, it harmonizes with data reported by other Authors for CLC in fixed beds (Mancuso et al., 2017). Main contribution to the levelized cost of energy arise from the price of the fuel and efficiency of the plant, responsible for about 65% of the total value of LCOE. A sensitivity analysis was performed by changing individually the cost of the different plant components. Fuel price is the main influencing variable, as a variation of ±15% in SFC induces a change in the LCOE of 10%. For the other components, when the cost changes within ±15%, the effect on LCOE is below 1.5%. 4. Conclusion The present study dealt with investigation of a two-stage chemical looping combustion process in fixed bed reactors and its related techno economic analysis. CH4 was considered as fuel, and Cu/CuO followed by Ni/NiO as oxygen carriers. Model results indicate an outlet gas temperature of about 1200 °C during power production, with an increase equally split between the two metal oxides. Analysis of the system transient operation disclosed that at least two parallel reactors are required for continuous power production. The reactor network was designed, showing that preheating of the reactants can be fulfilled by recovering sensible heat of the products. An overall power production of about 240 kWe was obtained, with an overall plant efficiency of about 22%. An investment of 424 k€ is estimated for the plant construction, mostly related to high temperature valves, compressors, and turbines. A levelized cost of energy of 134 €/MWhe is foreseen, in agreement with other similar technologies of chemical looping combustion in fixed beds. References Abad A., Adànez J., Garcìa-Labiano F., de Diego L.F., Gayàn P., Celaya J., 2007, Mapping the range of operational conditions for Cu-, Fe-, and Ni-based oxygen carriers in chemical-looping combustion, Chemical Engineering Science, 62, 533–549. Adànez J., Abad A., Garcìa-Labiano F., Gayàn P., de Diego L.F., 2012, Progress in Chemical-Looping Combustion and Reforming technologies, Progress in Energy and Combustion Science, 38, 215–282. Altimari P., Mancusi E., Russo L., Crescitelli S., 2012, Temperature wave-trains of periodically forced networks of catalytic reactors, AIChE Journal, 58, 899–913. Diglio G., Bareschino P., Mancusi E., Pepe F., 2017a, Sensitivity analysis in the design of a packed-bed reactor for a chemical looping combustion process, Chemical Engineering Transactions, 57, 1027–1032. Diglio G., Bareschino P., Solimene R., Mancusi E., Pepe F., Salatino P., 2017b, Numerical simulation of hydrogen production by chemical looping reforming in a dual fluidized bed reactor, Powder Technology, 316, 614–627. Diglio G., Bareschino P., Mancusi E., Pepe F., 2018, Techno-Economic Evaluation of a small-scale power generation unit based on a Chemical Looping Combustion Process in Fixed Bed Reactor network, Industrial & Engineering Chemistry Research, 57, 11299–11311. Garcia-Labìano F., de Diego L.F., Adànez J., Abad A., Gayàn P., 2004, Reduction and oxidation kinetics of copper-based oxygen carrier prepared by impregnation for chemical-looping combustion, Industrial & Engineering Chemistry Research, 43, 8168–8177. Hamers H.P., Romano M.C., Spallina V., Chiesa P., Gallucci F., van Sint Annaland M., 2015, Energy analysis of two stage packed-bed chemical looping combustion configurations for integrated gasification combined cycles, Energy, 85, 489–502. Iliuta I., Tahoces R., Patience G., 2010, Chemical-looping combustion process: Kinetics and mathematical modeling, AIChE Journal, 56, 1063–1079. Kooiman R. F., Hamers H.P., Gallucci F., Van Sint Annaland M., 2015, Experimental Demonstration of Two- Stage Packed Bed Chemical-Looping Combustion Using Syngas with CuO/Al2O3 and NiO/CaAl2O4 as Oxygen Carriers, Industrial & Engineering Chemistry Research, 54, 2001–2011. Mancusi E., Bareschino P. Forgione A., Pepe F., 2020, A two carriers reactor configuration for chemical- looping combustion in a packed-bed, International Journal of Greenhouse Gas Control, 99,103099. Mancuso, L., Cloete, S., Chiesa, P., Amini, S., 2017, Economic assessment of packed bed chemical looping combustion and suitable benchmarks, International Journal of Greenhouse Gas Control, 64, 223–233. Noorman S., van Sint Annaland M., Kuipers H., 2007, Packed Bed Reactor Technology for Chemical-Looping Combustion, Industrial & Engineering Chemistry Research, 46, 4212–4220. Tregambi C., Di Lauro F., Montagnaro F., Salatino P., Solimene R., 2019, 110th Anniversary: Calcium Looping Coupled with Concentrated Solar Power for Carbon Capture and Thermochemical Energy Storage, Industrial & Engineering Chemistry Research, 58, 21262–21272. Tregambi C., Bareschino P., Mancusi E., Pepe F., Montagnaro F., Solimene R., Salatino P., 2021, Modelling of a concentrated solar power – photovoltaics hybrid plant for carbon dioxide capture and utilization via calcium looping and methanation, Energy Conversion and Management, 230, 113792. 1008