CHEMICAL ENGINEERING TRANSACTIONS VOL. 81, 2020 A publication of The Italian Association of Chemical Engineering Online at www.cetjournal.it Guest Editors: Petar S. Varbanov, Qiuwang Wang, Min Zeng, Panos Seferlis, Ting Ma, Jiří J. Klemeš Copyright © 2020, AIDIC Servizi S.r.l. ISBN 978-88-95608-79-2; ISSN 2283-9216 Simulation of Carbon Capture from Flue Gas of a Coal-fired Power Plant by a Three-bed Pressure Swing Adsorption Process Chu-Yun Chenga, Chia-Chen Kuoa, Ming-Wei Yangb, Zong-Yu Zhuangb, Po-Wei Lina, Hong-Sung Yangc, Cheng-Tung Choua,* aDepartment of Chemical and Materials Engineering, National Central University, Zhongli District, Taoyuan City, Taiwan bChemistry and Environment Laboratory, Taiwan Power Research Institute, Shulin District, New Taipei City, Taiwan cDepartment of Applied Cosmetology, Hwa Hsia University of Technology, Zhonghe District, New Taipei City, Taiwan t310030@ncu.edu.tw In order to reduce carbon dioxide emissions, pressure swing adsorption (PSA) process was studied to capture carbon dioxide from flue gas in a coal-fired power plant. Pressure swing adsorption features its low energy consumption, low investment, and simple operation. This study aims to capture carbon dioxide from flue gas by PSA process for at least 85 % CO2 purity and with the other stream of more than 90 % N2 purity. To validate the accuracy of the PSA simulation program, the extended Langmuir-Freundlich equation was adopted to fit measured equilibrium data to describe the adsorption equilibrium of adsorbent zeolite 13X. Next, the simulation study used the linear driving force model and compared the results of breakthrough curves and desorption curves between experiments and simulation to verify the accuracy of the mass transfer coefficient kLDF value in linear driving force model. The agreement between experimental data and the simulation results is good. Further, the simulation was verified with the 100-hour cyclic-steady-state experiment of the 3-bed 9-step PSA process studied. Flue gas after desulphurisation and water removal (13.5 % CO2, 86.5 % N2) of subcritical 1 kW coal- fired power plant was taken as feed to the designed 3-bed 9-step PSA process. To find the optimal operating conditions, the central composite design (CCD) was used. After analysis, optimal operating conditions were obtained to produce a bottom product at 89.20 % CO2 purity with 88.20 % recovery, and a top product at 98.49 % N2 purity with 93.56 % recovery. The mechanical energy consumption was estimated to be 1.17 – 1.41 GJ/t-CO2. 1. Introduction According to the Global Carbon Budget climate report published by nearly 80 scientists (Le Quéré et al., 2018), the concerns of global carbon emission trends are pointed out, which show that slowing down carbon dioxide increase is urgent. There are three ways to capture carbon dioxide from power plants: post-combustion capture, pre-combustion capture, and oxyfuel combustion capture. The post-combustion capture technology is important due to its small impact on the industrial combustion process, so it is the main method in current carbon dioxide capture technology. For the gas separation, there are four main processes: Absorption, Cryogenic Separation, Membrane Separation, and Adsorption. Wang et al. (2013) compared different procedures, and the result has shown that the energy consumption of the adsorption method is better than the absorption method. Patil et al. (2018) used AMP/PZ/water mixture in a close-loop adsorber-desorber system to separate the CO2 from post- combustion gas (CO2-N2 mixture, CO2 12 mol %), resulting in 85 % removability and energy consumption is 3.78 MJ/kgCO2. Shen et al. (2012) used activated carbon in a two-stage vacuum pressure swing adsorption process to separate flue gas of 15 % CO2 as feed. The best experimental results can obtain 94.14 % CO2 purity with 85.08 % recovery, productivity 1.139·10-3 mol/kg-s, and the best simulation results were 96.34 % CO2 purity with 80.72 % recovery, productivity 2.58·10-4 mol/kg-s, and 829.28 kJ/kg-CO2 energy consumption. Wang et al. (2012) simulated a modified two-stage vacuum pressure swing adsorption program to capture carbon dioxide in the flue gas of 15 % CO2. The first stage used a 3-bed 5-step process, and the second stage uses a 2-bed DOI: 10.3303/CET2081070 Paper Received: 27/03/2020; Revised: 01/06/2020; Accepted: 07/06/2020 Please cite this article as: Cheng C.-Y., Kuo C.-C., Yang M.-W., Zhuang Z.-Y., Lin P.-W., Yang H.-S., Chou C.-T., 2020, Simulation of Carbon Capture from Flue Gas of a Coal-fired Power Plant by a Three-bed Pressure Swing Adsorption Process, Chemical Engineering Transactions, 81, 415-420 DOI:10.3303/CET2081070 415 6-step process. Both used zeolite 13X-APG as adsorbent and the results showed that 96.54 % CO2 purity with 93.35 % recovery, energy consumption 528.39 kJ/kgCO2, and productivity 8.61·10-6 kgCO2/kg-s were obtained. This study used EIKME 13X zeolite as the adsorbent and 13.5 % carbon dioxide balanced by nitrogen as feed. The feed flow rate of 1 kW coal-fired power plants referred to the National Energy Technology Laboratory (NETL) report (2015). In order to find the optimal operating conditions, this study used the central composite design for analysis. Before this study, there was no PSA process used for carbon capture in Taiwan’s power plant. The goal of this study is to develop a PSA process to capture CO2 for 1 kW flue gas in a coal-fired power plant of Taiwan Power Company, and the PSA process based on this paper is now operating at Taichung Power Plant. 2. Methodology 2.1 Mathematical modeling The assumptions made for mathematical modeling, the corresponding equations and boundary conditions are mentioned in Chou et al. (2013). The linear driving force model and extended Langmuir-Freundlich isotherm are used in the simulation. Extended Langmuir-Freundlich isotherm equation: q i *= ni * ρ s = q m,i biyi mi P mi 1+ ∑ bi n i=1 yi mi P mi (1) where q m,i =ai,1+(ai,2T), bi=bi,0+exp(bi,1/T), mi=mi,1+mi,2/T (2) In Eq(1), qi* represents the equilibrium adsorption amount of component i per unit adsorbent mass, ni* represents the equilibrium adsorption amount of component i per unit adsorbent volume, ρs is the density of the adsorbent, P is pressure, yi is the mole fraction of component i and T is temperature. The parameters used in Eq(2) are all isotherm parameters. Linear driving force model: ∂ni ∂t =kLDF,i( ni * - ni ) (3) In Eq(3), ni represents the adsorption amount of component i per unit adsorbent volume. 2.2 PSA process In this study, a 3-bed 9-step PSA process shown in Figure 1 was studied to capture CO2 by simulation, the process is described as follows: Adsorption (AD), Pressure Equalisation (PE), Cocurrent Depressursation (CD), Vacuum (VA) and Idle (ID), and EIKME 13X was used as the adsorbent. The feed composition of the process referred to Wawrzyńczak et al. (2019). The flue gas emitted by the power plant is first desulfurised, then removed the water resulting in 13.5 % carbon dioxide and 86.5 % nitrogen. The goal is to obtain higher than 85 % purity of carbon dioxide and 90 % purity of nitrogen in two different streams. In order to confirm the correctness of the parameters used in the program simulation, 100 h experimental data from a 3-bed 9-step 1 kW scale coal-fired power plant were used for verification. The procedure of the process is shown in Figure 1. The step time and the parameters of the bed are shown in Tables 1 and Table 2. Figure 1: Procedure of the 3-bed 9-step PSA process Table 1: Step time of 3-bed 9-step PSA process Step 1 2 3 4 5 6 7 8 9 Process AD AD PE CD VA PE ID ID AD Time (s) 80 300 50 80 300 50 80 300 50 416 Table 2: Parameters of the adsorption bed Parameter Value Feed flow rate (m3/s, NTP) 10-3 Bed length (m) 0.4 Bed diameter (m) 0.16 Feed temperature (K) 303.14 Feed pressure (atm) 3 Vacuum pressure (atm) 0.07 Cocurrent depressurisation pressure (atm) 0.25 3. Results and discussion 3.1 Adsorption isotherms The CO2 and N2 adsorption amount on EIKME 13X zeolite were measured by a micro-balance Thermo D-200. Langmuir-Freundlich isotherm, shown in Eq(1), was used as the equilibrium adsorption model and the parameters were fitted by The MATLAB Curve Fitting Toolbox and all parameters are shown in Table 3. The isotherm curves are shown in Figure 2 in which the error bars of 95 % confidence intervals are also shown. Figure 2: Adsorption isotherm of (a) CO2 and (b) N2 on EIKME 13X zeolite Table 3: Langmuir-Freundlich isotherm parameters of CO2 and N2 on zeolite 13X Carbon dioxide Nitrogen ai,1(mole/kg) 6.507 1.185·103 ai,2(mole/K·kg) -3·10-3 -1.022·10-1 bi,0(1/atm) 8·10-3 4.842·10-4 bi,1(K) 1.722·103 6.215·102 mi,1(-) 6.013·10-1 1.182 mi,2(K) -2.4·101 -1.2·102 3.2 Breakthrough curve and desorption curve verification To verify the reliability of the linear driving force coefficient and the simulation program, a breakthrough curve experiment was performed. Figure 3 shows that the result of simulation can fit the experimental data of breakthrough curve and desorption curve. The operating conditions are shown in Table 4. Table 4: Operating parameters of breakthrough curve and desorption curve Breakthrough experiment Desorption experiment Feed composition 15.0 % CO2 Pure He Bed length (m) 1 1 Bed diameter (m) 2.32·10-2 2.32·10-2 Bed volume (L) 0.42273 0.42273 Feed pressure (atm) 2.5 2.5 Feed temperature (K) 298 298 Surrounding temperature (K) 298 298 Feed flow rate (m3/s) 1.67·10-5 1.00·10-5 417 Figure 3: Simulation of (a) breakthrough curve and (b) desorption curves 3.3 3-bed 9-step PSA process verification In this section, the average experimental data of the last 40 cycles of the 3-bed 9-step 100-hour cyclic-steady- state experiment are used for simulation verification. The cyclic-steady-state feed concentration of CO2 is 15.36 % and concentration of N2 is 84.64 %, the feed flow rate is 9.703·10-4 m3/s, the feed pressure is 3 atm, the vacuum pressure is 0.07 atm, the vent pressure is 0.25 atm, the surrounding temperature is 303.14 K and the feed temperature is 303.14 K. Table 5: Results of the 3-bed 9-step PSA process Variables Experiment Simulation Top vent flow rate (m3/s, NTP) 4.77·10-5 7.03·10-5 CO2 Purity/ Recovery (%) 12.23/3.91 29.12/13.75 N2 Purity/ Recovery (%) - 70.88/6.08 Bottom product flow rate (m3/s, NTP) 8.92·10-5 8.92·10-5 CO2 Purity/ Recovery (%) 91.70/54.83 91.69/54.81 N2 Purity/ Recovery (%) - 8.31/0.90 Top product flow rate (m3/s, NTP) 8.05·10-4 8.11·10-4 CO2 Purity/ Recovery (%) 5.79/31.26 5.78/31.43 3.4 3-bed 9-step PSA process optimisation The central composite design was used to determine the optimal result. Variables are as below: feed pressure, vacuum pressure, co-current depressurisation pressure, ambient temperature, time of co-current depressurisation step, time of vacuum step, and tower length. The effects of these seven variables on the purity and recovery of carbon dioxide at the bottom product and energy consumption were discussed. Finally, regression analysis was used to find the optimal response and optimal operating conditions. The high- , basic- and low-level setting values of each factor are shown in Table 6. Eq(4), Eq(5), and Eq(6) are the regression equations for bottom product CO2 purity, bottom product CO2 recovery, and energy consumption, which can be used for optimisation. In Figure 4, the red line indicated the optimal results of the seven variables. The optimisation results using the setting value from Figure 4 are shown in Table 7 and Table 8. Table 6: Operating parameter Variables - 0 + A: Feed pressure (atm) 2.0 3.0 4.0 B: Vacuum pressure (atm) 0.05 0.075 0.1 C: Vent pressure (atm) 0.2 0.3 0.4 D: Surrounding temperature (K) 288.14 305.64 323.14 E: Step 1/4/7 (Cocurrent depressurisation) time (s) 40 80 120 F: Step 2/5/8 (vacuum) time (s) 250 300 350 G: Bed length (m) 0.3 0.4 0.5 𝑃𝑢𝑟𝑖𝑡𝑦 (%) = 208.683 + 8.44694 (1/𝑎𝑡𝑚) 𝐴 − 136.4 (1/𝑎𝑡𝑚) 𝐵 − 10.6067 (1/𝑎𝑡𝑚) 𝐶 − 1.03797 (1/𝐾) 𝐷 + 0.254009 (1/𝑠) 𝐸 + 0.0168206 (1/𝑠) 𝐹 + 0.753353 (1/𝑐𝑚) 𝐺 − 1.11434 (1/𝑎𝑡𝑚2) 𝐴 · 𝐴 + 303.611 (1/𝑎𝑡𝑚2) 𝐵 · 𝐵 − 14.7843 (1/𝑎𝑡𝑚3) 𝐶 · 𝐶 + 0.00176753 (1/𝐾²) 𝐷 · 𝐷 − 0.00158675 (1/𝑠2) 𝐸 · 𝐸 + 0.0001 (1/𝑠2) 𝐹 · 𝐹 + 0.000643 (1/𝑐𝑚2) 𝐺 · 𝐺 − 0.31307 (1/𝑎𝑡𝑚2) 𝐴 · 𝐵 + 12.9433 (1/𝑎𝑡𝑚2) 𝐴 · 𝐶 + 0.000388 (1/𝑎𝑡𝑚 · 𝐾) 𝐴 · 𝐷 + 0.015792 (1/𝑎𝑡𝑚 · 𝑠) 𝐴 · 𝐸 + 0.00384 (1/𝑎𝑡𝑚 · 𝑠) 𝐴 · 𝐹 − 0.08853 (1/𝑎𝑡𝑚 · 𝑐𝑚) 𝐴 · 𝐺 − 45.2696 (1/𝑎𝑡𝑚2) 𝐵 · 𝐶 + 0.436755 (1/𝑎𝑡𝑚 · 𝐾) 𝐵 · 𝐷 + 0.499657 (1/𝑎𝑡𝑚 · 𝑠) 𝐵 · 𝐸 − 0.239613 (1/𝑎𝑡𝑚 · 𝑠) 𝐵 · 𝐹 − 0.862748 (1/𝑎𝑡𝑚 · 𝑐𝑚) 𝐵 · 𝐺 − 0.0234598 (1/𝑎𝑡𝑚 · 𝐾) 𝐶 · 𝐷 − 0.323612 (1/𝑎𝑡𝑚 · 𝑠) 𝐶 · 𝐸 − 0.0206273 (1/𝑎𝑡𝑚 · 𝑠) 𝐶 · 𝐹 + 0.103386 (1/𝑎𝑡𝑚 · 𝑐𝑚) 𝐶 · 𝐺 + 0.0004568 (1/𝐾 · 𝑠) 𝐷 · 𝐸 − 4.61 · 10−5 (1/𝐾 · 𝑠) 𝐷 · 𝐹 − 0.00292673 (1/𝐾 · 𝑐𝑚) 𝐷 · 𝐺 − 1.59 · 10−4 (1/𝑠2) 𝐸 · 𝐹 + 0.00148013 (1/𝑠 · 𝑐𝑚) 𝐸 · 𝐺 − 2.7 · 10−4 (1/𝑠 · 𝑐𝑚) 𝐹 · 𝐺 (4) 418 𝑅𝑒𝑐𝑜𝑣𝑒𝑟𝑦 (%) = −54.8248 + 5.70294 (1/𝑎𝑡𝑚) 𝐴 − 290.665 (1/𝑎𝑡𝑚) 𝐵 − 1.79674 (1/𝑎𝑡𝑚) 𝐶 + 0.949833 (1/𝐾) 𝐷 − 0.100019 (1/𝑠) 𝐸 − 0.0570018 (1/𝑠) 𝐹 − 1.02886 (1/𝑐𝑚) 𝐺 − 1.816 (1/𝑎𝑡𝑚²) 𝐴 · 𝐴 − 1406.76 (1/𝑎𝑡𝑚²) 𝐵 · 𝐵 − 29.2673 (1/𝑎𝑡𝑚²) 𝐶 · 𝐶 − 0.0016378 (1/𝐾²) 𝐷 · 𝐷 − 1.65144 · 10−5 (1/𝑠²) 𝐸 · 𝐸 − 2.19 · 10−4 (1/𝑠²) 𝐹 · 𝐹 − 0.0115602 (1/𝑐𝑚2) 𝐺 · 𝐺 + 95.1987 (1/𝑎𝑡𝑚2) 𝐴 · 𝐵 − 4.91901 (1/𝑎𝑡𝑚²) 𝐴 · 𝐶 − 0.001048 (1/𝑎𝑡𝑚 · 𝐾) 𝐴 · 𝐷 − 0.0285889 (1/𝑎𝑡𝑚 · 𝑠) 𝐴 · 𝐸 − 0.00544839 (1/𝑎𝑡𝑚 · 𝑠) 𝐴 · 𝐹 + 0.186206 (1/𝑎𝑡𝑚 · 𝑐𝑚) 𝐴 · 𝐺 + 163.14 (1/𝑎𝑡𝑚2) 𝐵 · 𝐶 − 0.403525 (1/𝑎𝑡𝑚 · 𝐾) 𝐵 · 𝐷 − 0.9026 (1/𝑎𝑡𝑚 · 𝑠) 𝐵 · 𝐸 − 0.269244 (1/𝑎𝑡𝑚 · 𝑠) 𝐵 · 𝐹 + 3.48384 (1/𝑎𝑡𝑚 · 𝑐𝑚) 𝐵 · 𝐺 − 0.0252969 (1/𝑎𝑡𝑚 · 𝐾) 𝐶 · 𝐷 + 0.240571 (1/𝑎𝑡𝑚 · 𝑠) 𝐶 · 𝐸 + 0.0666479 (1/𝑎𝑡𝑚 · 𝑠) 𝐶 · 𝐹 + 0.135799 (1/𝑎𝑡𝑚 · 𝑐𝑚) 𝐶 · 𝐺 + 7.99 · 10−5 (1/𝐾 · 𝑠) 𝐷 · 𝐸 + 1.60842 · 10−4 (1/𝐾 · 𝑠) 𝐷 · 𝐹 + 0.00265063 (1/𝐾 · 𝑐𝑚) 𝐷 · 𝐺 − 1.21 · 10−4 (1/𝑠2) 𝐸 · 𝐹 + 0.0029009 (1/𝑠 · 𝑐𝑚) 𝐸 · 𝐺 + 0.0012076 (1/𝑠 · 𝑐𝑚) 𝐹 · 𝐺 (5) 𝐸𝑛𝑒𝑟𝑔𝑦 𝑐𝑜𝑛𝑠𝑢𝑚𝑝𝑡𝑖𝑜𝑛 (𝐺𝐽/𝑡 − 𝐶𝑂2) = 1.1382 + 0.714472 (1/𝑎𝑡𝑚) 𝐴 + 1.09506 (1/𝑎𝑡𝑚) 𝐵 + 0.448582 (1/𝑎𝑡𝑚) 𝐶 − 0.0099 (1/𝐾) 𝐷 + 0.00436957 (1/𝑠) 𝐸 − 0.00259758 (1/𝑠) 𝐹 + 0.0097060 (1/𝑐𝑚) 𝐺 − 0.038474 (1/𝑎𝑡𝑚²) 𝐴 · 𝐴 + 36.4096 (1/𝑎𝑡𝑚²) 𝐵 · 𝐵 + 0.0355262 (1/𝑎𝑡𝑚²) 𝐶 · 𝐶 + 1.85 · 10−5 (1/𝐾²) 𝐷 · 𝐷 + 6.29 · 10−6 (1/𝑠²) 𝐸 · 𝐸 + 6.12 · 10−6 (1/𝑠²) 𝐹 · 𝐹 + 1.85442 · 10−4 (1/𝑐𝑚²) 𝐺 · 𝐺 − 0.0859038 (1/𝑎𝑡𝑚²) 𝐴 · 𝐵 − 0.00750255 (1/𝑎𝑡𝑚²) 𝐴 · 𝐶 − 2.83 · 10−4 (1/𝑎𝑡𝑚 · 𝐾) 𝐴 · 𝐷 + 0.00132517 (1/𝑎𝑡𝑚 · 𝑠) 𝐴 · 𝐸 − 2.33 · 10−4 (1/𝑎𝑡𝑚 · 𝑠) 𝐴 · 𝐹 − 0.00373936 (1/𝑎𝑡𝑚 · 𝑐𝑚) 𝐴 · 𝐺 − 4.36119 (1/𝑎𝑡𝑚²) 𝐵 · 𝐶 − 0.00479232 (1/𝑎𝑡𝑚 · 𝐾) 𝐵 · 𝐷 + 0.032285 (1/𝑎𝑡𝑚 · 𝑠) 𝐵 · 𝐸 + 0.0011352 (1/𝑎𝑡𝑚 · 𝑠) 𝐵 · 𝐹 − 0.08924 (1/𝑎𝑡𝑚 · 𝑐𝑚) 𝐵 · 𝐺 + 7.96313 · 10−4 (1/𝑎𝑡𝑚 · 𝐾) 𝐶 · 𝐷 − 0.005784 (1/𝑎𝑡𝑚 · 𝑠) 𝐶 · 𝐸 − 1.21 · 10−4 (1/𝑎𝑡𝑚 · 𝑠) 𝐶 · 𝐹 + 0.00448796 (1/𝑎𝑡𝑚 · 𝑐𝑚) 𝐶 · 𝐺 − 5.44 · 10−6 (1/𝐾 · 𝑠) 𝐷 · 𝐸 − 1.27 · 10−6 (1/𝐾 · 𝑠) 𝐷 · 𝐹 − 1.23 · 10−5 (1/𝐾 · 𝑐𝑚) 𝐷 · 𝐺 − 7.19 · 10−6 (1/𝑠²) 𝐸 · 𝐹 − 6.56 · 10−5 (1/𝑠 · 𝑐𝑚) 𝐸 · 𝐺 − 1.20 · 10−5 (1/𝑠 · 𝑐𝑚) 𝐹 · 𝐺 (6) Figure 4: The optimal response of regression from purity, recovery and energy consumption Table 7: Optimal operating condition of capturing CO2 from flue gas by PSA process Parameter Value Feed composition 13.5 % CO2, 86.5 % N2 Feed flow rate (m3/s, NTP) 1.21·10-3 Bed length (m) 0.46 Bed inner diameter (m) 0.16 Bed volume (L) 9.2488 Bed porosity (-) 0.6937 Fluid viscosity (kg/m-s) 1.87·10-5 Overall heat transfer coefficient (J/m2 ·s·K) 10.8 Feed temperature (K) 303.14 Surrounding temperature (K) 323.14 Feed pressure (atm) 3.66 Vacuum pressure (atm) 0.05 Cocurrent depressurisation pressure (atm) 0.3 Step time (s) 94, 350, 50, 94, 350, 50, 94, 350, 50 Cur H igh Low0.68039 D O ptim al d = 0.60292 M axim um Purity y = 90.0730 d = 0.58182 M axim um Recovery y = 89.5456 d = 0.89791 M inim um Energy y = 1.1531 0.68039 D esirability Com posite 30.0 50.0 250.0 350.0 40.0 120.0 288.140 323.140 0.20 0.40 0.050 0.10 2.0 4.0 B C D E F GA [3.6566] [0.050] [0.2909] [323.140] [94.1414] [350.0] [45.7576] 419 Table 8: The results of simulation before and after central composite design analysis from purity, recovery and energy consumption Variables Simulation for basic case Prediction from Regression Simulation after CCD Purity (%) 85.96 89.83 89.20 Recovery (%) 82.09 89.78 88.20 Energy consumption (GJ/t) 1.06 1.15 1.17 4. Conclusions In this study, the isotherm parameters were obtained by fitting the CO2 and N2 adsorption data from the experiment. The breakthrough curve and desorption experiments performed at the lab-scale were used to verify the simulation accuracy. The agreement is good. The 3-bed 9-step experimental process was used to capture the flue gas from the power plant for the 100 h cyclic-steady-state experiment. The average composition of the feed gas is 15.36 % carbon dioxide and 84.64 % nitrogen for the last 40 cycles of the 100-hour PSA experiment. The comparison between the experiments and simulation show that the simulation result is reliable. Finally, this study used the central composite design (CCD) and regression analysis to find the optimal purity, recovery, and minimum energy consumption with 13.5 % carbon dioxide 86.5 % nitrogen feed composition. The result showed that the purity of the carbon dioxide in the bottom product can reach 89.20 % with a recovery 88.20 %, and the purity of the nitrogen in the top product can reach 98.49 % with a recovery 93.56 %, and the power consumption was 1.17 - 1.41 GJ/t-CO2 under optimal conditions: feed pressure = 3.66 atm, vacuum pressure = 0.05 atm, co- current depressurisation pressure = 0.3 atm, surrounding temperature = 323.14 K, step 1/4/7 time = 94 s, step 2/5/8 time = 350 s and tower length = 0.46 m. Acknowledgements The authors wish to thank the Taiwan Power Company for financial support. References Chou C., Chen F., Huang Y., Yang H., 2013, Carbon dioxide capture and hydrogen purification from synthesis gas by pressure swing adsorption, Chemical Engineering Transactions, 32, 1855 - 1860. Le Quéré C., Andrew R.M., Friedlingstein P., Sitch S., Hauck J., Pongratz J., Pickers P.A., Korsbakken J.I., Peters G. P., Canadell J.G., Arneth A., Arora V.K., Barbero L., Bastos A., Bopp, L., Chevallier F., Chini L. P., Ciais P., Doney S.C., Gkritzalis T., Goll D.S., Harris I., Haverd V., Hoffman F. M., Hoppema M., Houghton R.A., Hurtt G., Ilyina T., Jain A.K., Johannessen T., Jones C.D., Kato E., Keeling R.F., Goldewijk K.K., Landschützer P., Lefèvre N., Lienert S., Liu Z., Lombardozzi D., Metzl N., Munro D.R., Nabel J.E.M.S., Nakaoka S., Neill C., Olsen A., Ono T., Patra P., Peregon A., Peters W., Peylin P., Pfeil B., Pierrot D., Poulter B., Rehder G., Resplandy L., Robertson E., Rocher M., Rödenbeck C., Schuster U., Schwinger J., Séférian R., Skjelvan I., Steinhoff T., Sutton A., Tans P.P., Tian H., Tilbrook B., Tubiello F. N., van der Laan-Luijkx I.T., van der Werf G.R., Viovy N., Walker A.P., Wiltshire A.J., Wright R., Zaehle S., Zheng B., 2018, Global carbon budget 2018, Earth System Science Data, 10, 2141 – 2194. National Energy Technology Laboratory, 2015, Cost and Performance Baseline for Fossil Energy Plants, The United State Department of Energy, 1a, rev. 3. Patil M., Vaidya P., Kenig E., 2018, Bench-scale study for CO2 capture using AMP/PZ/water mixtures, Chemical Engineering Transactions, 69, 163 - 168. Shen C.Z., Liu Z., Li P., Yu J.G., 2012, Two-stage VPSA process for CO2 capture from flue gas using activated carbon beads, Industrial & Engineering Chemistry Research, 51, 5011 - 5021. Wang L., Liu Z., Li P., Wang J., Yu J.G., 2012, CO2 capture from flue gas by two successive VPSA units using 13XAPG, Adsorption Journal of the International Adsorption Society, 18, 445 – 459. Wang L., Yang Y., Shen W., Kong X., Li P., Yu J., Rodrigues A.E., 2013, CO2 capture from flue gas in an existing coal-fired power plant by two successive pilot-scale VPSA units, Industrial & Engineering Chemistry Research, 52, 7947 – 7955. Wawrzynczak D., Majchrzak-Kuceba I., Srokosz K., Kozak M., Nowak W., Zdeb J., Smolka W., Zajchowski A., 2019, The pilot dual-reflux vacuum pressure swing adsorption unit for CO2 capture from flue gas, Separation and Purification Technology, 209, 560 – 570. 420