Electric swing adsorption as emerging CO2 capture technique - Core

Available online at www.sciencedirect.com

Energy Procedia

EnergyProcedia Procedia1 00 (2008) 000–000 Energy (2009) 1219–1225

www.elsevier.com/locate/procedia www.elsevier.com/locate/XXX

GHGT-9

Electric swing adsorption as emerging CO2 capture technique Carlos A. Grande*, Rui P. L. Ribeiro, Eduardo L. G. Oliveira, Alirio E. Rodrigues LSRE – Laboratory of Separation and Reaction Engineering. Associate Laboratory. Faculdade de Engenharia, University of Porto. Rua Dr. Roberto Frias (4200-465) Porto, PORTUGAL Elsevier use only: Received date here; revised date here; accepted date here

Abstract The application of Electric swing adsorption (ESA) as a successful second generation technique to capture CO2 was evaluated. The objective of this work is to demonstrate that under certain conditions, ESA can be used in CO2 capture from flue gases of power plants. A process with seven steps was developed and tested for CO2 capture from a flue gas from a natural gas power station with emission levels of 1 Gton of CO2 per year with a content of 3.5%. The seven steps employed are: feed, rinse with hot gas, internal rinse, electrification, depressurization and two purge steps. Using this process we could obtain CO2-rich streams with purities of 89.7% keeping recovery around 70% and energy consumption of 1.9 GJ/ton of CO2 captured. c 2008

2009 Elsevier © Elsevier Ltd. Ltd.Open All rights access reserved under CC BY-NC-ND license. Keywords: Electric Swing Adsorption; CO2 capture; natural gas

1. Introduction The global emissions of anthropogenic CO2 and other greenhouse gases (GHG) have increased significantly in the last years enhancing the natural greenhouse effect and producing the so-called global warming [1]. Power generation is one of the major stationary sources of emissions of CO2. Power stations using different fossil fuels (coal, oil and natural gas) have ensured secure energy supply and direct availability. If we want to continue using these resources we should find a strategy to remove one of the combustion products (CO2) and avoid its emissions to atmosphere. One of the solutions is to store the captured CO2 in secure geological locations, although its use in a new CO2-based chemistry can also be considered. The main problem of this solution relies on the technology employed to capture CO2. Flue gases of current power plants are a mixture of nitrogen, oxygen, carbon dioxide and water plus other minor contaminants. Capture CO2 from these streams is termed as post-combustion capture since the CO2 removal takes place after the power generation. A successful technology to capture CO2 from these streams can be retrofitted to existing power plants, enhancing the short-term reduction of GHG emissions. There are no commercial large scale power stations operating yet with post-combustion capture. Even though, fast advances are taking place all over the world, both in adapting and scaling up known CO2 capture techniques and also on finding secure geological locations for storage. Several techniques to capture CO2 in post-combustion power

* Corresponding author. Tel.: +351-22-508-1618; fax: +351-22-508-1618. E-mail address: [email protected].

doi:10.1016/j.egypro.2009.01.160

1220

2

C.A. Grande et al. / Energy Procedia 1 (2009) 1219–1225 Grande et al./ Energy Procedia 00 (2008) 000–000

plants are under scope. The most studied one is amine scrubbing [2-5], while scrubbing with other solvents [6], membranes [7-9] and Pressure Swing Adsorption [10-12] are also being fostered. The main focus of these techniques was in coal-fired flue gases where the content of CO2 is around 15%. On the other side, several technological problems were found when the flue gas comes from a natural gas fired power stations. In this case, the gas contains only 3-5% of CO2 and large amounts of oxygen. Also, increasing the purity of CO2 from 3.5% to 95% means an enrichment factor of 27 which is very difficult to achieve without spending large amounts of energy. In this case, adsorption technologies can provide a large and selective surface where the separation can take place efficiently. The main objective of this work is to present the Electric swing adsorption (ESA) technology as an innovative second generation technique to capture CO2 from flue gases. Particular interest will be given to treat flue gases of natural gas fired power plants with low CO2 content. In this initial process design, we have considered that the flue gas is composed by 3.5% of CO2 balanced by nitrogen. ESA is a cyclic and sequential process composed by different steps to use and regenerate the adsorbent in an efficient way. By now, there is no specific recipe to design and to arrange the steps in ESA processes and more than one configuration can lead to similar results. In this work we show a new cyclic configuration able to achieve a high purity CO2 stream and also with some energetic integration to reduce the power consumption (penalties) of the process. 2. ESA process for CO2 capture Electric Swing Adsorption is the name that was given to an adsorption process where the regeneration is performed by increasing the temperature of the adsorbent using the Joule effect of passing electricity through a conductor [13-16]. In fact, regeneration of the adsorbent is carried out by the reducing the equilibrium capacity of the materials. In these terms, it seems that the principles of operation of the ESA process are similar with the operation of the most known Temperature Swing Adsorption (TSA), but there are substantial differences in unit productivities (electric heating is much faster) and to concentrate non-condensable gases (heating is not performed with a diluent). The adsorbent employed in ESA experiences large differences in temperature, so significant differences in loading at low and high temperatures are expected. This is normally translated to high values of heats of adsorption and maximum performance can be achieved if the concentration of the gas to be adsorbed is small. This process is not suitable for removal of large contents of gas; in those cases Pressure Swing Adsorption seems more appropriate [17]. The ESA process is mentioned in several reports as one possible technique to capture CO2 from flue gases, but so far this process was commercially employed to remove volatile organic compounds [18-23]. Only two technical works report some experimental data on CO2 capture, both using commercial adsorbents. We have employed a honeycomb monolith of activated carbon [24] while the group of Prof. Grevillot has used a bed filled with zeolite 13X extrudates [25]. We have ensured good electrical properties, but activated carbons have very low CO2 loadings at low partial pressures making impossible its direct use for this purpose. On the other side, zeolite 13X ensures high capacity at the expense of requires external wiring to conduct electricity which may be rather difficult to scale-up to treat flue gases from a power station. An adsorbent with good electrical properties (able to conduct electricity) and with high loading of CO2 at low partial pressures is not available. In order to determine the applicability of the technology for post-combustion capture we have assumed that the adsorbent is available and is constituted by a composite material comprising 80% of zeolite 13X crystals merged by 20% of a conducting matrix that can be graphite for example [26]. Using this adsorbent, we have simulated a five-step ESA cycle. The cycle comprised a feed, an internal rinse where the column started to be heated passing electricity but allowing some gas to leave by the top, electrification (where the high temperature is achieved by passing electricity in the closed column), depressurization and purge. A CO2-purified stream is collected both in the depressurization and purge steps. It was also determined that for an inlet flue gas with 3.5% of CO2, if the adsorbent has a capacity higher than 1.05 mol/kg, a cooling step is not required and cooling is done co-currently with the feed step. With this five-step cycle we were able to recover 79.5% of the CO2 in the flue gas with a purity of 79.4%. The energy required was 2.04 GJ/ton of CO2 captured. At least 90% of the energy was required to heat the adsorbent from 310K to 455K while the rest of the energy is used in the fan required to recycle 5-7% of the purified stream to purge the columns.


1221 3

With this five-step cycle we have obtained an enrichment of 22.6 times. This value can be considered as very good, but unfortunately is not enough if we are interested in producing a CO2 stream with 95% purity. To achieve higher purities, we have two different alternatives. We can think about a material having a higher capacity, or we can change the arrangement of steps to improve the ESA performance. Materials with higher capacity can be obtained by combining the same binder with novel adsorbents that can achieve loadings of 2.94 mol/kg at 5 kPa of CO2 [27]. This capacity is significantly higher than the values we have assumed for our adsorbent and thus much higher performance is possible. It should be pointed out that an adsorbent with very high capacity is highly preferred since less material is used and thus less energy is used for heating. In this work we have not changed the properties of the adsorbent but made some modifications to the cycle presented previously achieving higher purities and also reducing the energy penalties. Our mathematical simulations are based on a previously published model used to describe the separation of CO2 by Electric Swing Adsorption (ESA) using an activated carbon honeycomb monolith [24]. This model considers that the solid is bidisperse (composed by micropores and macropores), that the pressure drop is described by the Darcy equation and also with energy balances in the gas and solid phases as well as in the column wall. The set of partial differential equations was solved in gPROMS software. The complete list of operating conditions employed in the simulations is detailed in Table 1. The simulations were performed in order to treat a flue gas from a gas fired power station with emissions of 1 Gton of CO2 per year and with a content of CO2 of 3.5% (balanced by N2). The existence of oxygen is not quite relevant since adsorption properties of N2 and O2 in zeolite 13X is quite similar. The most important simplification made here is the fact that water was previously removed. We have considered that the adsorbent is composed by 80% of zeolite 13X which provides all the capacity for adsorption [28] while the other 20% do not contribute to adsorb CO2 and only provides a conducting media for electricity. We have assumed that the adsorption kinetics in zeolite 13X crystals is not affected and thus the crystal diffusivities are according to values reported in literature [29]. The adsorption equilibrium data was described using the multi-site Langmuir model (parameters detailed in Table 1). Table1. Conditions employed in ESA simulation using the seven-step cycle developed Column and feed properties Step Times Column length (m) 11.62 tfeed (s) 600 Column diameter (m) 7.38 thot rinse (s) 180 Column porosity 0.40 tinternal rinse (s) 480 Feed flowrate (m3/s) 330.87 telectrification (s) 900 Feed temperature (K) 310.15 tdepressurization (s) 60 CO2 molar fraction 0.0350 tpurge (s) 42 Adsorbent Heat capacity (J/mol.kg) 900 tpurge to recycle (s) 180 Wall density (kg/m3) 1166.7 tcycle (s) 2442 Adsorption Equilibrium Parameters CO2 N2 qmax (mol/kg) 3.00 3.00 (-¨H) (kJ/mol) 45.0 20.0 K0 (kPa-1) 4.25u10-9 4.6u10-7 ai 1.0 1.0 Adsorption kinetic parameters Ea (kJ/mol) 22.0 25.1 Dc0 (m2/s) 5.5u10-11 4.6u10-10 3. Results and Discussion The main reason of the purities around 80% obtained in our previous work, was because the CO2 was recovered from two different streams: the exit of depressurization and purge steps. The stream leaving the depressurization step has normally purity higher than 95%, but the stream leaving the purge is highly contaminated with nitrogen (purity around 70%).

1222

4


In order to improve the process performance, we have decided to modify the ESA cycle by introducing new steps to include the possibility of recycling some CO2. At the end of the feed step, the column has the CO2 adsorbed and a large amount of moles of N2 in the gas phase (96.5% in the entire column). Afterwards, some of this adsorbed CO2 is desorbed by heating the column, displacing the N2 from the gas phase, consuming electricity. Other important observation noted before is that if we decrease the duration of the purge step, the decrease in recovery is drastic: achieving purities around 90% will lead to recoveries smaller than 50%. It should be pointed out that the stream exiting the purge step is at high temperatures. Merging these observations, it seems that a partial recycle of the stream leaving the purge step may fitfully work here. Recycling a hot stream with a higher CO2 content will help reducing the N2 content in the column prior to electrification and also (once that the stream is hot) we will be able to reduce the electricity consumption once we are pre-heating the column for desorption. As the content of CO2 in the purge step decreases exponentially with time, the purge can be divided into two steps. In the first one, the gas with high contents of CO2 is still recovered as product while in the second step, the stream exiting is recycled. One extreme is total recycle of the purge and the other one is total recovery, which is the option that was previously employed [26]. The resulting new scheme is shown in Figure 1. In the image, seven steps are employed: feed, rinse with recycled hot gas, internal rinse to deploy nitrogen from the column, electrification, depressurization, purge and the final step that is the purge to provide gas for the recycle.

Figure 1. Scheme of the different steps employed in the ESA process to capture CO2 from flue gases. Steps are: (1) feed; (2) rinse with recycled gas; (3) internal rinse; (4) electrification; (5) depressurization; (6) purge; (7) purge to provide gas for recycle. To have a direct comparison with previous results without rinsing with heavy gas, we have kept the properties of the column and its dimensions as well as composition and flowrate of the feed step (see Table 1). A very interesting property of the ESA cycles is the easiness to reach the cyclic steady state (CSS). After the first cycle, almost all variables are very close to the CSS value, which is attained after four cycles. An example of the results obtained with this new cycle is shown in Figure 2. The mass transfer zone within the column in the feed step is quite spread, although very little diffusional problems exist. The heat of adsorption of CO2 released in the feed step results in a temperature increase of 20K (Fig. 2d) that has a very strong influence in the CO2 adsorption capacity (Fig. 2b), spreading the mass transfer zone. The new rinse step has an important role in increasing the content of CO2 in the column before the thermal regeneration, both in gas and adsorbed phases. Important conclusions from this example can be taken from the molar flowrates exiting the column (Fig. 2c). It can be observed that in the feed step there is a fast pulse of CO2 that corresponds to all CO2 not removed in the purge steps. The other important thing to be observed is the decrease of CO2 in the last steps of the cycle (6 and 7). In this simulation, the purity obtained was 89.8% with 72.0% of recovery. It can be observed that the purge step (step 6) has some steady contamination with CO2 but the amount of CO2 strongly decreases. The overall purity of this stream is 80.0% that mixed with a purity of 96.5% coming from the depressurization step, results in the overall value of 89.8%. Thus, reducing the time of

1223


5

purge, higher values of purity can be obtained, but there is a penalty in the recovery: if more gas is recycled to the column, it will break through the adsorbent. Larger columns are not an interest solution since more power should be employed to heat them. Regarding the power consumption, using the recycle of hot gas, it was possible to reduce 4 minutes of application of electricity to attain the same temperature at the end of the electrification step. On the other side, as we are including a new recycle in the cycle, a new fan should be introduced to recompress the stream exiting the purge to a higher pressure to deal with the pressure drop. In the example provided in this work, we have a power consumption of 1.9 GJ/ton of CO2 captured. This value is not very different from our previous value because we are also losing more CO2 (recovery of 72%). 2

(a) 5

0.8 0.6

6

Amount adsorbed [mol CO2/ kg]

CO2 molar fraction

1 4

3

0.4 2

0.2 7

1

0 2

4 6 8 Column length [m]

10

1.6 1.4

1

1.2 1

3

0.8 5

0.6 0.4 7

0

10000

2000

8000

0

5

6

2150 2250 2350 2450

6000 1

4000

2

3

4

2000

12

5

4

440 420

7

400

6

380

3

360

2

320

CO2 0

10

(d)

340

0


460

7

Temperature [K]

4000

2

480

(c) N2

12000

4 6

0.2

12

14000

(b)

2

0 0

Molar flowrate [mol/ s]

1.8

1

300 500

1000

1500

Cycle time [seconds]

2000

2500

0

2


10

12

Figure 2. Simulation results of an ESA process for CO2 capture from flue gas of natural-gas fired power plant: a) CO2 molar fraction; b) CO2 amount adsorbed; c) molar flowrate; d) temperature. The values are reported at the end of each step in cyclic steady state. Conditions employed are detailed in Table 1. Steps are: (1) feed, 600s; (2) rinse, 180s; (3) internal rinse, 480s; (4) electrification, 900s; (5) depressurization, 60s; (6) purge, 42s; (7) purge for recycle, 180s. This example show that the ESA technique has a power consumption that is comparable with the values obtained for CO2 capture from flue gases of coal-fired power stations, where the content of CO2 is around 15%. This is an indicative that ESA technique can be competitive with other capture techniques. To have this technique ready for operation, several technical issues should be addressed. In particular, the feasibility of having an adsorbent with the adsorption characteristics described in this work, without serious kinetic limitations for adsorption in the zeolite 13X crystals. Other situation that should be solved is the influence of water. Zeolites adsorb water very strongly, reason

1224

6


why water should be previously removed. This can be done even with activated carbon that can be used as an initial layer or in a separate pre-treatment unit. Alternatively, the adsorbent can be made of other material that can withstand water and perform some chemical bonding with CO2 that can then be regenerated with electricity. In this case, the electric consumption can be reduced significantly. 4. Conclusions We have studied the performance of Electric Swing Adsorption as merging technique for CO2 capture from flue gases of natural gas power stations where the CO2 content is 3.5%. A new configuration with a partial recycle of CO2 was evaluated as a pre-heating technique that can also help us to enrich the concentration of this gas prior to its desorption. The cycle employed comprise seven steps: feed, rinse with hot CO2-rich stream, internal rinse, electrification, depressurization and purge that was divided into two steps. The CO2-rich stream is recovered in the depressurization step and in the initial moments of the purge. With this new cycle configuration we have achieved a CO2 rich stream with purity of 89.7% and recovering 72.0% of the CO2 of the inlet flue gas. The energy consumption of the process was estimated to be 1.9 GJ/ton of CO2 captured, which is in the same order as other technologies employed in CO2 capture from flue gases with 15% of CO2. The results obtained in this work provide strong evidence that ESA can be one of the technologies that deserve further studies as second-generation capture technique. 5. References 1. Metz, B., Ogunlade, D., Connink, H., Loos, M., Meyer, L., 2005. Special Report on Carbon Dioxide Capture and Storage. Special Report of the Intergovernmental Panel on Climate Change. 2. Hook, R. J., 1997. An Investigation of Some Sterically Hindered Amines as Potential Carbon Dioxide Scrubbing Compounds. Ind. Eng. Chem. Res. 36, 1779 -1790. 3. Yeh, J. T., Pennline, H. W., Resnik, K. P., 2001. Study of CO2 Absorption and Desorption in a Packed Column. Energy & Fuels, 15, 274-278. 4. Alie, C., Backham, L., Croiset, E., Douglas, P.L. 2005. Simulation of CO2 capture using MEA scrubbing: a flowsheet decomposition method. Energy Conversion and Management, 46 (3), 475-487. 5. Romeo, L.M., Bolea, I., Escosa, J., 2008. Integration of power plant and amine scrubbing to reduce CO2 capture costs. Applied Thermal Engineering, 28 (8-9), 1039-1046. 6. Zeman, F., 2007. Energy and material balance of CO2 capture from ambient air. Environmental Science and Technology, 41 (21), 7558-7563. 7. Bredesen, R., Jordal, K., Bolland, O., 2004. High-temperature membranes in power generation with CO2 capture. Chemical Engineering and Processing: Process Intensification, 43 (9), Membrane Reactors, 1129-1158. 8. Favre, E., Bounaceur, R., Lape, N., Roizard, D., Vallieres, C., 2006. Membrane processes for post-combustion carbon dioxide capture: A parametric study. Energy, 31 (14), 2220-34. 9. Chiesa, P., Kreutz, T.G., Lozza, G.G., 2007. CO2 sequestration from IGCC power plants by means of metallic membranes. Transactions of the ASME. Journal of Engineering for Gas Turbines and Power, 129(1),123-34. 10. Reynolds, S. P., Ebner, A. D., Ritter, J. A., 2005. New pressure swing adsorption cycles for carbon dioxide sequestration. Adsorption, 11 (1) Suppl., 531-536. 11. Ho, M. T. Allinson, G. W., Wiley, D. E., 2008. Reducing the cost of CO2 capture from flue gases using pressure swing adsorption, Ind. Eng. Chem. Res., 47 (14), 4883-4890. 12. Zhang, J., Webley, P. A., 2008. Cycle development and design for CO2 capture from flue gas by vacuum swing adsorption. Environmental Science and Technology, 42 (2), 563 – 569. 13. Petkovska, M., Tondeur, D., Grevillot, G., Granger, J., Mitrovic, M., 1991. Temperature-Swing Gas Separation with Electrothermal Desorption Step. Sep. Sci. Technol., 26, 425-444. 14. Burchell, T.D., Judkins, R.R., Rogers, M.R., Williams, A.M., 1997. A novel process and material for the separation of carbon dioxide and hydrogen sulfide gas mixtures. Carbon 35, 1279–1294. 15. Saysset, S., Grevillot, G., Lamine, A.S., 1999. Adsorption of Volatile Organic Compounds on Carbonaceous Adsorbent and Desorption by Direct Joule Effect. Recent Progress in Genie des Procedes, Nº.68. Proceedings of the 2nd European Congress of Chemical Engineering, Montpellier, France, 389–396.


1225 7

16. Sullivan, P.D., 2003. Organic vapor recovery using activated carbon fiber cloth and electrothermal desorption. Ph.D. Thesis. University of Illinois at Urbana. 17. Ruthven, D.M., 1984. Principles of Adsorption and Adsorption Processes, Wiley & Sons, New York. 18. Place, R.N.; Blackburn, A.J.; Tennison, S.R., Rawlinson, A.P.; Crittenden, B.D., 2005. Method and Equipment for Removing Volatile Compounds from Air. US patent 6,964,695. 19. Judkins, R.R.; Burchell, T.D., 1999. Electrical swing adsorption gas storage and delivery system, US Patent 5,912,424. 20. Farant, J.P.; Desbiens, G., 2008. Adsorption of contaminants from gaseous stream and in situ regeneration of sorbent, US patent 7,316,731. 21. Sullivan, P.D., Rood, M. J. Grevillot, G., Wander, J. D., Hay, K. J., 2004. Activated carbon fiber cloth electrothermal swing adsorption system Environ. Sci. Technol. 38, 4865-4877 22. Yu, F.D., Luo, L., Grevillot, G., 2006. Electrothermal swing adsorption of toluene on an activated carbon monolith. Experiments and parametric theoretical study. Chem. Eng. Process. 46, 70–81. 23. Luo, L., Ramirez, D., Rood, M. J., Grevillot, G., Hay, K. J., Thurston, D. L., 2006. Adsorption and electrothermal desorption of organic vapors using activated carbon adsorbents with novel morphologies. Carbon 44, 2715 – 2723. 24. Grande, C. A.; Rodrigues, A. E. 2008. Electric Swing Adsorption for CO2 Removal from Flue Gases. Int. J. Green. Gas Control, 2, 194-202. 25. Ettlili, N., Bertelle, S., Roizard, D., Vallieres, C., Grevillot, G., A New Electrical Swing Adsorption Process for Post-Combustion CO2 Capture, Proceedings of the 8th International Conference on Greenhouse Gas Control Technologies, Elsevier, 2006. ISBN: 0-08-046407-6. 26. Grande, C. A., Ribeiro, R. P. L., Rodrigues, A. E., (2008) Electric swing adsorption for CO2 removal. 2nd Petrobras International Seminar on CO2 Capture and Geological Storage. Salvador, Brazil (2008). 27. Harlick, P. J. E., Sayari, A., 2007. Applications of Pore-Expanded Mesoporous Silica. 5 .Triamine Grafted Material with Exceptional CO2 Dynamic and Equilibrium Adsorption Performance. Ind. Eng. Chem. Res., 46, 446458. 28. Cavenati, S.; Grande, C. A. and Rodrigues, A. E., 2004. Adsorption Equilibrium of Methane, Carbon Dioxide and Nitrogen on Zeolite 13X at High Pressures. J. Chem. Eng. Data, 49, 1095-1101. 29. Cavenati, S., Grande, C. A., Rodrigues, A. E., 2006. Removal of Carbon Dioxide from Natural Gas by Vacuum Swing Adsorption. Energy & Fuels, 20, 2648-2659.