On the development of Vacuum Swing adsorption (VSA) technology ...

Available online at www.sciencedirect.com

Energy Procedia 37 (2013) 33 – 39

GHGT-11

On the development of Vacuum Swing adsorption (VSA) technology for post-combustion CO2 capture Anne Andersena, Swapnil Divekarb, Soumen Dasguptab,Jasmina Hafizovic Cavkaa, Aartib, Anshu Nanotib,Aud Spjelkavika, Amar N. Goswamib, M.O.Gargb and Richard Bloma* a

SINTEF Materials & Chemistry, P.O.Box 124 Blindern, 0314 Oslo, Norway CSIR-Indian Institute of Petroleum, Dehradun 248005, Uttrakhand, India

b

Abstract A metal-organic framework, UiO-66, has been evaluated as adsorbent in a post-combustion vacuum swing adsorption (VSA) process. Equilibrium isotherms of the most relevant gases (CO2 and N2) as well as breakthrough curves measured using synthetic flue gas containing 15 mol% CO2 without and with 9 mol% water vapor are reported. Based on the breakthrough data, a six step one-column VSA cycle is designed and the effects of adsorption and CO2 rinse times used on the CO2 recovery and CO2 purity are examined. With the chosen process configuration and cycle design CO2 purities around 60% and CO2 recoveries up to 70% are achieved. 50 cycle adsorption-desorption experiments show that the cyclic CO2 capacity is reduced by approximately 25% in the presence of water vapor. No reduction in cyclic capacity is observed with increased cycle number; there is rather a slight increase in cyclic capacity with cycle number indicating that a cyclic steady state still not has been reached after 50 cycles. © 2013 The Authors. © Authors Published Publishedby byElsevier ElsevierLtd. Ltd. Selection and/or of of GHGT Selection and/or peer-review peer-reviewunder underresponsibility responsibility GHGT Keywords: metal-organic framework, MOF, UiO-66, CO2 capture; flue gas; adsorption; VSA

1. Introduction Rising levels of CO2 in the atmosphere due to burning of fossil fuel have been recognized to be the main contributor of global warming and associated climate change phenomenon. Fossil fuel combustion

* Corresponding author. Tel.: +47 90622647; fax: +47 22067350. E-mail address: [email protected].

1876-6102 © 2013 The Authors. Published by Elsevier Ltd.

Selection and/or peer-review under responsibility of GHGT doi:10.1016/j.egypro.2013.05.082

34

Anne Andersen et al. / Energy Procedia 37 (2013) 33 – 39

for power generation is the major source of increased CO2 levels in the atmosphere, but capturing CO2 from flue gas emissions in power plants where it is available as a low pressure stream, presents a formidable challenge. Current technologies available such as amine based absorption processes are considered uneconomic and there is a concerted effort being made to improve such processes or develop more efficient alternative processes. In this context adsorption processes using Pressure or Vacuum Swing Adsorption (PSA/VSA) are attracting interest as energy requirements are lower [1][2]. Solid adsorbents like zeolites and activated carbons can be used to recover CO 2 from flue gas mixtures by pressure swing adsorption technique. Several adsorbent materials have been investigated for CO2 recovery by PSA/VSA. The general consensus appears to be that Zeolite 13X materials performs better than activated carbons or silica gels for low pressure applications [3][4]. Both capacities and selectivities for separation of CO2/N2 mixtures (representative of flue gases from power plants) are superior. However, as CO2 isotherms on zeolites are nonlinear due to its relatively high adsorption energy (>40 kJ/mole), power requirement during regeneration can be high and there is for this reason a large scope for developing new adsorbents which will show better selectivity and regenerability. Metal-organic frameworks (MOF) is a new class of adsorbents attracting interest for selective CO 2 separation [5]. These are materials in which metal ions or clusters are connected via organic linkers to 2

recovery, however, the several studies that have been reported so far on CO2 been limited mostly to equilibrium isotherm and diffusion measurements with pure components [6]. Recently a comparative study of VSA performance of a commercially available MOF (CuBTC) with a commercial 13X zeolite has been reported for CO2 recovery from CO2-N2 gas mixture [7]. The present paper reports an experimental study on the separation of CO 2 from mixtures with nitrogen using UiO-66 (Zr6O4(OH)4(1,4-dicarboxybenzene)) [ 8 ]adsorbent in a vacuum swing adsorber(VSA) operating with a heavy reflux or rinse cycle typically used for recovery of the strong adsorptive (in this case CO2) from gas mixtures. CO2 adsorption on nano-particles UiO-66 has recently been reported [9], however, for engineering purposes the in-house synthesized UiO-66 presented in the present paper has been formulated in suitable forms for column operation. Basic adsorption characteristics as well as results from multi-cycle experiments using multicomponent gas mixtures including moisture are reported. Based on the experimental data a VSA cycle is developed. 2. Experimental 2.1. Adsorbent preparation and basic characterization UiO-66 was prepared following the original procedure used by Cavka et al. [8]. Spherical adsorbent beads of about 1.7 mm were prepared by the following method: 1.0 g UiO-66 was suspended in 9.0 g of a 0.5wt% alginate-water solution. The slurry was stirred for 2 hrs.at ambient temperature. The slurry was the dropwise added to a 0,034 m Ba(NO3)2. The beads formed was washed twice with water, and then dried at room temperature overnight. A specific surface area (BET) of 1260 m2/g was measured after activation in vacuum at 130 ºC overnight. The bulk density of the bed of spheres was approximately 0.45g/ml. N2 and CO2 isotherms of the formulated adsorbent were measured at a volumetric Belsorp Max instrument. The isotherms were recorded from vacuum to 1 atm. at 303 and 328K. The isotherms are shown in Figure 1 together with lines indicating Langmuir fits to the experimental points. 2.2. Breakthrough and Single Column VSA experiments The breakthrough and Vacuum Swing Adsorption experiments were carried out in a fully automated single column breakthrough unit described in reference [6]. The unit consists of an adsorber column of 11


mm I.D. and 105 mm packed length. Solenoid valves are used for directing gas flows to and from the column. Binary feed gas mixtures of required composition of CO 2 in nitrogen were prepared from pure gas cylinders using mass flow controllers. Separate mass flow controllers were used for flow control of purge (nitrogen) and rinse (CO2) streams, respectively. Column pressure is controlled by a Pressure Control Valve. A PC based SCADA was used for data acquisition (temperature, pressure, flows) and operation of solenoid valves. CO2 concentration in the various gas streams were measured by IR-based on-line gas analysis. In all breakthrough experiments, a feed gas mixture containing 15% CO2 in nitrogen representing a flue gas from a coal -fired power plant was used. During a typical breakthrough experiment the adsorber column was maintained at constant temperature under flow and pressure control while the CO2 concentration in the effluent was analysed at constant time interval (every 10th second) until the CO2 concentration of the effluent reached that of the feed gas. Multi-cycle adsorption-desorption experiments were carried out with both dry and wet feed gas at temperature 55 ºC and pressure 2 bar respectively to study the cyclic stability of the adsorbent for CO2 capture. To mimic the most realistic wet flue gas, 9 mole % water vapour was introduced into the gas mixture by using a high pressure pump (Eldex Pump). The timings for adsorption and regeneration steps are fixed at 900 and 1620 sec respectively. The VSA cycle studies have been carried out at 2 bar pressure and 55 °C temperature with a dry feed mixture of 15% CO2 in balance nitrogen at fixed feed flow rate of 0.26 NL/min. The VSA cycle involved six consecutive steps namely 1.Feed Pressurisation 2.Adsorption 3.Counter-current blow-down, 4.Cocurrent rinse with CO2, 5.Counter-current Evacuation, and 6.Counter-current Evacuation with Nitrogen purge as described in reference [7]. Feed pressurisation, blow-down and evacuation with nitrogen purge time were fixed at 20 sec, 5 sec and 18 sec respectively. Other cycle step times were chosen based on the CO2 column breakthrough time observed: The adsorption step time were chosen to vary from 41% to 61% with respect to the CO2 breakthrough time. For a particular adsorption time the evacuation time were similarly varied from 41% to 61% of the CO2 breakthrough time. For a particular adsorption and evacuation time the co-current rinse time was varied from 9% to 18% of CO2 breakthrough time. The sequence of different steps and their timings are described in Table 1. Table 1: Description of PVSA Cycle Step No. Step description 1.

Duration

Co-current Pressurization

20 sec (Fixed)

2.

Adsorption

41% /51% / 61% of CO2 Breakthrough Time (BT)

3.

Co-current- Blow-down

5 sec (Fixed)

4.

Co-current CO2 Rinse

9%/ 13.6 %/ 18% of CO2 BT

5.

Counter-current Evacuation

41% /51% /61% of CO2 BT

6.

Evacuation + Nitrogen Purge

18 sec (Fixed)

VSA Cycle Illustrative Case for Description: CO2 breakthrough time with 15% CO2 balance N2 feed @ 2 bar pressure, 55 ºC temperature and 0.26 NL/min feed flow rate: Breakthrough time (BT)= 283 sec. VSA Step Times: 1. Co-current Pressurization: 20 sec; 2. Adsorption: 173 sec ( 61% of CO2 BT); 3. Co-current- Blow-down: 5 sec; 4. Co-current CO2 Rinse: 25 sec ( 9% of CO2 BT); 5. Counter-current Evacuation: 116 sec ( 41% of CO2 BT); 6. Evacuation + Nitrogen Purge: 18 sec Total VSA Cycle Time:357 sec

Pure gas cylinders were used for sourcing the rinse and purge streams. Around 25 cycles were performed in each experiment to allow cyclic steady state to be achieved before measuring CO 2 concentrations in each stream from the adsorber.

35

36


3. Results and discussion The pure component adsorption isotherms of CO2 and Nitrogen on the UiO-66 adsorbent are shown in Figure 1 for pressures up to 1 bar. As expected, equilibrium CO2adsorption capacity decreases with increase in temperature. However, at lower pressure range (up to 1 bar), the shape of the isotherm tends to become more linear with increase in temperature.

Figure 1: Pure component adsorption isotherms of CO2 and Nitrogen on the UiO-66 adsorbent. The lines indicate fitted Langmuir models.

Breakthrough curves with feed mixture of CO2 and N2 at three pressures (2, 3 and 4 bar) are reported in Figure 2. As can be seen, the breakthrough time increases as pressure increases. This is because the adsorbent capacity for CO2 increases as pressure increases as indicated by the CO2 isotherms shown in Figure 1. The breakthrough curves are sharp suggesting good mass transfer rates in UiO-66.

Figure 2: Breakthrough curves at different pressures with feed mixture of 15 mole% CO2 balance N2


CO2 breakthrough time of 283 sec at 2 bar was used to design the single column VSA experiments. The VSA cycle step times chosen were based on the CO2 column breakthrough time at a particular pressure, temperature, feed flow rate and feed composition as described in Table 1. The effect of variation of adsorption step time on the purity and recovery of CO2 is shown in Figure 3. With increasing adsorption step time, a gradual increase in the purity and recovery of CO2 in the evacuation product was observed. This observation may be ascribed to more penetration of the mass transfer front into the adsorbent bed and more CO2 adsorption with increase in adsorption time. The effect of CO2 rinse time is also shown in Figure 3. As the CO2 rinse time is increased the void fluid is displaced by pure CO2 and hence the purity of the CO2 obtained in the subsequent evacuation step improves. The recovery values also show increasing trend with increase in rinse time.

Figure 3: Effect of variation of adsorption time (left) and CO2 rinse time (right) on the CO2 purity (light grey bars) and CO2recovery (dark grey bars).

In order to check the stability of UiO-66 during multi-cycle operation, 50 consecutive adsorptiondesorption cycles were carried out with both dry and moist feed. Cyclic breakthrough experiments with moist feed were carried with a feed mixture comprising of 15 mole% CO2, 9 mol% water vapour and balance N2 at 55 ºC and 2 bara pressure. Figure 4 shows the representative breakthrough and regeneration curves with moist feed. The 50 curves nearly overlap; in fact there is a slight trend towards longer breakthrough times for increasing cycle number that is not clearly visible in the present graph.

Figure 4: Variation of breakthrough time for successive cycles for moist feed

37

38


To obtain a clearer view on any trend in the breakthrough times over the 50 cycles we have plotted the variation in breakthrough times for the 50 successive cycles carried out with both dry and moist feed in Figure 5: Firstly, there is a reduction of around 25% in the CO2 breakthrough time in the presence of moisture compared to the dry feed. The reduction is steady over the 50 cycles and no further deterioration in the performance of the adsorbent in subsequent cycles is observed. The results indicate that the reduction in cyclic CO2 adsorption capacity in the presence of H2O is a consequence of adsorptive competition between the two species and no accumulation of H2O on the adsorbent seems to occur at the conditions used since this should lead to reduced CO2 breakthrough time with increased cycle number. Secondly, there is a slight increase in breakthrough time with cycle number that indicate that a true steady state situation has not been reached during the first 50 cycles and that more cycles are needed to reach a steady state performance. This is the case for both the dry and moist case.

Figure 5: Variation of breakthrough time (BT) for successive cycle for dry and moist feed

4. Conclusions We have studied a metal-organic framework, UiO-66, during prolonged adsorption-desorption cycling both at dry and moist conditions using a synthetic flue gas containing 15 mol% CO 2 as feed at 55ºC. The results show that the cyclic CO2 capacity is reduced by around 25% when water vapour is introduced, however, no loss in cyclic capacity with cycle number is observed indicating that the adsorbent is chemically stable at the conditions used. Actually, there is a steady increase in cyclic CO2 capacity over the 50 cycles used indicating that cyclic steady state still has not been reached. The breakthrough data has further been used to design a six step VSA cycle: CO2 recovery between 60 and 70% and CO2 purities around 60% is obtained with the simple cycle. The purity and recovery values are expected to improve further in a multi-bed VSA set up with scaled up adsorbents and improved VSA cycle design. Acknowledgements This work has been carried out with financial support from The Norwegian Ministry of Foreign Affairs in New Delhi (contract IND 3025 09/059)


References [1] Ho MT, Allinson GW, Wiley DE.Reducing the cost of CO2 capture from flue gases using pressure swing adsorption.Ind. Eng. Chem. Res. 2008;47;4883-4890. [2] Liu Z, Wang L, Kong X, Li P, Yu J, Rodrigues AE. Onsite CO 2 capture from flue gas by an adsorption process in a coalfired power plant. Ind. Eng. Chem. Res. 2012; 51;7355-7363. [3] Li G, Xiao P, Webley P, Zhang J, Singh R, Marshall M. Capture of CO 2 from flue gas by vacuum swing adsorption with zeolite 13X. Adsorption. 2008;14;415-422. [4] Chue KT, Kim JN, Yoo YJ, Cho SH, Yang RT. Comparison of activated carbon and zeolite 13X for CO2 recovery from flue gas by pressure swing adsorption. Ind. Eng. Chem. Res. 1995;34;591-598. [5]See recent reviews: Samantha A, Zhao A, Shimizu GKH, Sarkar P, Gupta R. Post-combustion CO2 capture using solid sorbents: A review. Ind. Eng. Chem. Res. 2012;51;1438-1463; Li J-R, Ma Y, McCarthy MC, Sculley J, Yu J, Jeong H-K, Balbuena PB, Zhou H-C. Carbon dioxide capture-related gas adsorption and separation in metal-organic frameworks. Coord. Chem. Rev. 2011;255;1791-1823; Sumida K, Rogow DL, Mason JA, McDonald TM, Bloch ED, Herm ZR, Bae T-H, Long JR. Carbon dioxide capture in metal-organic frameworks. Chem. Rev. 2012;112;724-781.. [6] See for example: Choudhary P, Bikkina C, Meister D, Dreisbach F, Gumma S. Comparison of adsorption isotherms on CuBTC metal organic frameworks synthesised from different routes. Micro.Meso.Mater. 2009;117;406-413; Cavenati S, Grande CA, Rodrigues AE, Kiener C, Müller U. Metal organic framework adsorbent for biogas upgrading. Ind. Eng. Chem. Res. 2008;47;63336335; Liang Z, Marshall M, Chaffee AL. CO2 adsorption-based separation by metal organic framework (CuBTC) versus Zeolite (13X). Energy& Fuels 2009;23;2785-2789; Hamon L, Jolimaitre E, Pirngruber GD. CO2 and CH4 separation by adsorption using CuBTC metal organic framework. Ind. Eng. Chem. Res. 2010;49;7497-7503; Krishna R, Long, JR. Screening metal-organic frameworks by analysis of transient breakthrough of gas mixtures in a fixed bed adsorber. J. Phys. Chem. C. 2011;115;1294112950; Schell J, Casas N, Blom R, Spjelkavik AI, Andersen A, Cavka JH, Mazotti M. MCM-41, MOF and UiO-67/MCM-41 adsorbents for pre-combustion CO2 capture by PSA: Adsorption equilibria. Adsorption 2012;18;213-227. [7] Dasgupta S, Biswas N, Arya A, Gode N, Divekar S, Nanoti A, Goswami AN. CO2 recovery from mixtures with nitrogen in a vacuum swing adsorber using metal organic framework adsorbent: A comparative study. Int. J. of Greenhouse Gas Control 2012;7;225 229 [8] Cavka J.H , Jakobsen S, Olsbye U, Guillou N, Lamberti C, Bordiga S, and Lillerud KP. A new zirconium inorganic building brick forming metal organic frameworks with exceptional stability.J. Am. Chem. Soc.2008; 130; 13850 13851 [9] Abid HR, Tian H, Ang H-M, Tade MO, Buckley CE, Wang S. Nanosize Zr-metal organic framework (UiO-66) for hydrogen and carbon dioxide storage. Chem. Eng. J. 2012;187;415-420.

39