Study on Carbon Dioxide Removal from Flue Gas by Absorption of Aqueous Ammonia

Qiao He, Minggong Chen, Lingyu Meng, Kunlei Liu* and Wei-Ping Pan Institute for Combustion Science and Environmental Technology Western Kentucky University, Bowling Green, KY, 42101 Introduction It has been known for more than 100 years that carbon dioxide (CO2) is a greenhouse gas and the release of CO2 from fossil fuel may affect the climate of the earth.1 The growing awareness of the risks of climate changes has generated public concerns and, since 1989, the interests of researchers in sequestering CO2. As the world population increases and energy demand rises, increased burning of fossil fuels will continue to drive atmospheric CO2 levels upward. The Inter-government Panel on Climate Change (IPCC) predicts that atmospheric CO2 concentrations will be increased by approximately twice from the pre-industrial concentration of 270 ppm to a concentration of 530 ppm in 2050 and could potentially exceed 700 ppm by 2100.2 This increase will significantly affect the global weather and the physiological basis of plant production unless major changes are made. According to the United Nations Framework Convention on Climate Change (UNFCCC) at Kyoto,3 a commitment to reduce CO2 emissions by 6% below the 1990s level was made by several countries. President Bush also announced the Global Climate Change Initiative (GCCI) with the goal of significantly reducing the U.S. greenhouse gas intensity by 18% by 20124. Power generation contributes one-third of the CO2 released from fossil fuel combustion worldwide,5 which cause future generation of CO2 through combustion will be substantial. The goal of reducing CO2 concentrations in the atmosphere can be met either by decreasing the rate at which CO2 is emitted into the atmosphere or by increasing the rate at which it is removed from it.6 A lot of research has been focused on determining a fast and inexpensive method to sequester carbon nowadays. There are two typical strategies to reduce CO2 emission from power plant combustion, one is to concentrate the CO2 in-situ during the fuel conversion process, and the other one is to extract the CO2 from the combustion flue gas followed by sequestration process. Since the fuel conversion process is usually very complex and difficult to control, the former technique is still highly investigative using.7 Several possibilities for the latter method have been proposed and developed, such as chemical solvents, physical absorption, cryogenic methods, membrane systems, biological fixation, and so on.8-10 Physical solvents are favored by high pressures and low concentrations of inert gases. CO2 can be physically absorbed in a nonreactive solvent according to Henrys law and then regenerated using pressure reduction or heat. Chemical solvents are preferred for cases with low concentrations or amounts of CO2 in the combustion gases and do not gain significant advantage by operating at elevated pressure. The chemical solvent methods are generally recognized as the most effective technologies at present. However, the cost for separation of CO2 is high, typically in the range $40-200 /ton of carbon.11 In addition, the chemical solvent method also has several major problems including a slow absorption rate, small solvent capacity, and special equipment requirements. To reduce the cost of capturing CO2, a low cost solvent that can minimize the operation cost, the equipment size, and equipment corrosion is desirable. Among a lot of absorbents, amines have been considered as typical CO2 absorbents. Chosen for principal commercial efforts in CO2 capture and regeneration, they include monoethanolamine (MEA), diethanolamine (DEA) and methyldiethanolamine (MDEA). In addition to aqueous solutions of alkanolamines, proprietary formulations comprising mixtures of the amines with various additives are also widely used. The MEA process has been thought as one of the most successful amine CO2 absorption methods. In the MEA process, the MEA aqueous solution absorbs CO2 at 100oF or slightly higher, and when the CO2-rich MEA solution is heated to approximately 250oF, the CO2 gas is released in

high purity. The MEA process is termed regenerable because the CO2 is absorbed at low temperature and then released at a higher temperature. The regenerated, high purity CO2 gas stream could be sequestered by a variety of technologies, and the regenerated MEA solution that remains is in a CO2-lean condition, which can be reused for CO2 capture. Thus, the MEA process has been widely used for CO2 regeneration. However, it is an expensive option with the cost of CO2 separation in the range from $40 to $70/ton of CO2 removed due to its low loading capability, slow absorption rate and so on. It is not a best way to control large amount of CO2 emission from utility flue gas. A current commercial approach in fertilizer industry which may provide an inexpensive and effective route of reducing CO2 emissions from power plants is extracting CO2 by an ammonia (NH3) reagent in a wet scrubber. A wet pressurized scrubber is used to produce NH4HCO3 currently. How to de-pressurize is a challenge to us. Injection of NH3 gas or aqueous NH3 for removing NOx (selective catalytic reduction) from flue gas is a common process in power plants. Unlike the Monoethanolamine (MEA) process, the aqueous ammonia process does not have absorbent degradation problems, which are caused by sulfur dioxide and oxygen in the flue gas and does not cause equipment corrosion, as in the case of MEA12. In the work by Bai and Yeh,13 based on breakthrough curves, the NH3 scrubbing capacity was calculated to be around 0.35 mol of CO2/mol of NH3 on a molar basis, or 0.9 -1.2 kg of CO2/kg of NH3 on a mass basis. The removal efficiency is approximate 99%. Yeh And Bai14 concluded that the maximum CO2 removal efficiency by NH3 absorbent can reach 99% and the CO2 loading capacity can approach 1.20 Kg CO2/Kg NH3. However, the maximum CO2 removal efficiency and loading capacity by MEA absorbent were 94% and 0.40 Kg CO2/Kg MEA, respectively. In the work by Smouse,15 a multi-pollution control concept with spraying aqueous ammonia into actual flue gas to capture CO2, SO2 and NOX emissions was proposed and developed. The capture efficiency of CO2 in absorber was in the range from 76.4% to 91.7% at 35 oC. The products in the aqueous ammonia scrubber could include ammonium bicarbonate, ammonium carbonate, and ammonium carbamate (all in crystalline or aqueous solution forms), plus reagent CO2, NH3, and NH4OH. Of these by-products, ammonium bicarbonate has been utilized by certain developing countries as a crop fertilizer for over 30 years with proven results. A process using coal char to produce ammonium bicarbonate for fertilizer has been developed and used commercially over 50 years. As an alternative to use ammonium bicarbonate as fertilizer, the ammonia (NH3) in ammonium bicarbonate can be regenerated. Ammonium bicarbonate decomposes at the relatively low temperature of 60oC, compared to a 120oC regeneration temperature for MEA solutions. The preliminary experiment of Bai and Yeh results led to a conceptual solvent regenerable process where the crystalline NH4HCO3 product is heated to regenerate the NH3.13 With its lower costs, higher loading capacity and higher CO2 absorption efficiency, lower decomposition temperatures, less corrosive environment for absorber material and good performance of fertilizer, the ammonium scrubbing technique presents many advantages for the capture of CO2 over the conventional MEA process. Theory of NH3-CO2-H2O reaction Although the aqua ammonia scrubbing process is an attractive way to capture low concentration CO2 from flue gas, chemistry reaction among CO2, ammonium and H2O is very complex and poorly understood many species co-exist in the solution in an unstable transition. These species could include, but not limited to, ammonium bicarbonate, ammonium carbonate, and ammonium carbamate (all in crystal and aqueous solution forms), plus CO2, NH3, and NH4OH. To meet the need for a CO2 capture process, the reactions must be investigated in detail, along with the composition of flue gas, the reaction rate, CO2 partial pressure and solubility, product separation, and so on. CO2 can be removed by ammonium scrubbing through chemical absorption at various temperatures and operating conditions. As illustrated by equation (1), ammonium carbamate (NH2COONH4) is the main product in dry conditions (without moisture) under ambient conditions. However, ammonium carbamate, is very soluble in water. Under moist air, the hydration product of ammonium carbonate, (NH4)2CO3, is produced at room temperature, as illustrated in equation (2).13

CO2 ( g ) + 2 NH 3 ( g ) NH 2 COONH 4 ( s ) (NH3 /CO2 molar ratio = 0.5)

(1)

(3) The gas-liquid chemical reactions between NH3 and CO2 in the wet scrubber can be illustrated by the following equations:13 (4) CO2 ( g ) + 2 NH 3 ( g ) CO( NH 2 ) 2 ( s) + H 2 O( g ) (NH3/CO2 molar ratio = 0.5)
+ CO2 ( g ) + 2 NH 3 (l ) NH 4 (l ) + NH 2 COO (l )

NH 2 COONH 4 ( s ) + H 2 O( g ) ( NH 4 ) 2 CO3 ( s ) NH 2COONH 4 ( s ) + H 2O(l ) NH 4 HCO3 ( s ) + NH 3 ( g )

(2)

2 NH 3 ( g ) + CO2 ( g ) + H 2 O( g ) ( NH 4 ) 2 CO3 ( s)

(NH3/CO2 molar ratio = 0.5) (NH3/CO2 molar ratio = 0.5)

(5) (6) (7) (8) (9)

(NH3/CO2 molar ratio = 1.0) NH 3 ( g ) + CO2 ( g ) + H 2 O( g ) NH 4HCO3 ( s) 2 NH 3 (l ) + CO2 ( g ) + H 2 O(l ) ( NH 4 ) 2 CO3 ( s ) (NH3/CO2 molar ratio = 0.5) (NH3/CO2 molar ratio = 1.0) NH 3 (l ) + CO2 ( g ) + H 2 O(l ) NH 4HCO3 ( s)

At room temperature and atmospheric pressure, reactions (4)-(9) would occur. These reactions are very temperature sensitive. Due to the different concentrations of reactants and reaction conditions (temperature and pressure), different carbon-ammonium composites (i.e., different ammonium salts), can be obtained, as illustrated by equation (10):
aNH 3 + bCO 2 + cH 2 O NH 4 HCO3 (C salt ); ( NH 4 ) 2 CO3 H 2 O( S salt );

2 NH 4 HCO3 ( NH 4 ) 2 CO3 ( P salt );

NH 4 COONH 2 ( A salt )

(10)

Equation (11) indicated that ammonium bicarbonate and ammonium carbonate could be converted to each other at certain condition,
( NH 4 ) 2 CO3 ( s ) + H 2 O(l ) + CO 2 ( g ) 2 NH 4 HCO3 ( s )

(11) All the products are white solids and may be a single salt or mixed salts. Reaction (8) and (9) are the most probable in this study for CO2 removal by NH3 scrubbing. In view of achieving a maximum of NH3 utilization in the capture of CO2 (minimum inventory required for CO2 sequestration), reaction (9) will be ideal. Theoretically, a maximum 2.59 kg of CO2/kg of NH3 can be reached if only reaction (9) occurs in the wet scrubber. In commercial processes, the reaction follows the following two steps: gaseous CO2 dissolves in solution and reacts with aqueous ammonia. Based on previous studies, the reaction process is diffusion controlled dissolution of gaseous CO2 in solution. Therefore, increasing CO2 dissolubility in water (solution) could enhance the reaction rate. Experimental Study The schematic diagram of the experimental system for studying on removing CO2 gas by ammonium scrubbing is shown in Figure 1. The CO2 scrubber was a glass container (I.D. 50mm) filling with 150 ml of 15% ammonium solution. The CO2 gas was fed from a simulated flue gas cylinder with CO2 concentration 14.7%. A mass flow controller is used to control the inlet CO2 flow rate. A filter was placed after the scrubber to capture the escaped particles. Photoacoustic multi-gas analyzer (INNOVA 1312) was used for monitoring the outlet CO2 and ammonium concentration. An magnet stirring system is used to help the reactant vertically mixed well. Experiments were conducted at room temperature condition (251C). Flow rate of flue gas was kept as 3 L/min. The ammonium solution was obtained from a standard ammonium solution (Fisher Science, 29.7% (w/w)), which was diluted using DI water (in some case spent-solution was used as the thinner). In order to investigate the influence of additional MEA on CO2s ammonium scrubbing, a few milliliter of MEA solution was additionally used with ammonium solution together. Some experiments were also used spent solution from previous experiments as the thinner to dilute the standard ammonium solution instead of DI water, or dissolving ammonium bicarbonate into the ammonium solution to make the initial ammonium solution in a certain carbonate degree. The influence of such additives on reaction efficiency was discussed in this paper. Results from seven experiments have been presented in this paper. Six of them were run at the stirring system under

different operating condition. One experiment was operated without stirring. The operation conditions for all seven experiments are listed in the Table 1.

Multi-gas analyzer Outlet gas

Bypass

Filter MFC

Ammonium solution

Temperature monitor Stirring

Syngas

Reactor

Figure 1. Schematic diagram of experiment setup Table 1. Operation Conditions

Results and Discussion Experiments 1 6 have been conducted in a system with magnet stirring. The variation of CO2 concentrations as a function of operation time has been shown in Figure 2. It has been found from the results, that CO2 outlet concentration dropped into a very low level (below 1ppm) at beginning of all experiments. The outlet concentrations of CO2 could be maintained at the low level for about 30 min. With the reaction going on, the CO2 outlet concentrations increased gradually, and finally approached to the CO2 inlet concentration, 14.7%. This indicates that the ammonium solution has already reached its saturated capability to absorb CO2. In order to evaluate the capacity of the ammonium reagent, all experiments were lasted until the NH3 scrubbing capacity became zero. The longest operation time was up to 6 hrs. CO2 outlet concentrations have been demonstrated in Figure 2. Those curves acted as breakthrough curves for the NH3 reagent. The calculation and analysis from DSC and TGA showed that both ammonium bicarbonate and ammonium carbonate could appear in the solution at the final stage of the reactions. It may be explained that (NH4)2CO3 was generated first at the solution (equation 8). When the CO2 continuously bobbled into system, (NH4)2CO3 might absorb CO2 to form NH4HCO3, as shown in equation (11).

CO2 Outlet Concentration (ppm)

16 12 8 4 0 0: 00

1: 12

2: 24

3: 36

4: 48

6: 00

7: 12

time
Exp 1 Exp 2 Exp 3 Exp 4 Exp 5 Exp 6

Figure 2. Variation of CO2 outlet concentration under different operation condition Effect of Stirring on CO2 Removal Reaction
20 15 10 5 0 0: 00: 00

CO2 outlet concentration (ppm)

1: 12: 00

2: 24: 00

3: 36: 00

4: 48: 00

time Exp 1 Exp 7

Figure 3. Effect of stirring on CO2 removal reaction Effect of stirring on CO2 removal is shown in Figure 3. Result from Experiment 1 and 7 are plotted in Figure 3. Exp 7 was the system without stirring. After 2 hrs of the reaction, the CO2 outlet concentration approached to the 14%, and NH3 solution did deteriorate at all. However, a slightly shaking of the reactor bottle caused the CO2 outlet concentration dropped immediately. This phenomenon was not found in the stirring system, where the CO2 outlet concentrations were gradually approaching to the CO2 inlet concentration and lasted at this level even the time of CO2 supply was extended to 1 hr longer. This indicated that ammonium solution was already fully reacted with CO2, and did not have more absorption capability. The reaction system with stirring could lead to a continuous and gradual reaction process, which is good for ammonium to play its CO2 absorption role. A system without stirring will probably indicate a fake signal that ammonium solution is degraded, however, there are still a lot of effective ammonium available at solution. This experiment showed that it is important to have a mixing or turbulent effect on the reaction, no matter for the open experiment system or future continuous reaction system. Influence of additional MEA on CO2 Removal Efficiency The MEA has been thought as a successful CO2 absorbent in practically industrial application except its high operating cost. It does have a good affinity to CO2, and the reaction between MEA and CO2 is more stable than the inorganic ammonium process to absorb CO2. An interest was paid on the MEA-ammonium mixed reagent. As part of the reagent, 1.5ml MEA has been added into reaction system (Exp 3), where

MEA contained 1% of the reagent solution. Results from Exp 1 and Exp 3 have been compared in Figure 4. It was interested to noted that the CO2 removal efficiency was typically higher for the MEA added system than the regular ammonium scrubbing system. At the beginning of the reaction, the CO2 absorption efficiency could reach as high as 99%. And it did keep the higher absorption efficiency through all the reaction time. Since the CO2 affinity ability for MEA is stronger than ammonium, the MEA is reacted with CO2 more rapidly. Certain amount of CO2 was absorbed by MEA, while the ammonium scrubbing was take action afterwards. The adding of MEA did not dramatically improve the ammonium CO2 absorption capability. It did like two absorption processes integrated together and there was a time delay between Exp 1 and Exp 3 in CO2 outlet concentration curves. The difference of CO2 removal efficiency in Exp 1 and Exp.3 was due to the contribution of additional MEA.

CO2 removal efficiency (%)

100 80 60 40 20 0

0:00:00

1:12:00 time
Exp 1

2:24:00

3:36:00

Exp 3

Figure 4. Influence of additional MEA on CO2 removal process Influence of reuse of spent solution on CO2 removal efficiency

CO2 removal efficiency (%)

100 80 60 40 20 0 0:00:00 1:12:00

2:24:00

3:36:00

Figure 5. Influence of reuse of spent-solution on CO2 removal efficiency The effective ammonium concentration at solution is a critical factor to influence the CO2 absorption rate. The higher its concentration is, the faster CO2 absorption rate is, and greater the absorption capability. Experiment results from Exp 1, Exp 2 and Exp 4 were compared in Figure 5. CO2 with concentration 14.7% was bobbled to 15% ammonium solution (Exp 1). 5 ml (Exp 2) and 30 ml (Exp 4) spent-solution were added into solution as part of the reagent. Figure 5 demonstrated that the CO2 removal efficiency

was increased largely and the removal process was more stable when more spent-solution was added as reagent. The results also implied that a continuous system would have higher CO2 removal efficiency than the incontinuous system since the spent-solution was always recirculated into reactor in the continuous system. Influence of adding additional NH4HCO3 on CO2 removal efficiency

CO2 removal efficiency (%)

100 80 60 40 20 0 0: 00
Exp 1

1: 12
time

2: 24
Exp 5

3: 36
Exp 6

Figure 6. The variation of CO2 removal efficiency vs. time Similar with spent-solution, additional NH4HCO3 crystal dissolved into reaction solution can also help to enhance the carbonated degree of reagent solution, which will help the ammonium solution to increase its capability of CO2 removal. Three operating conditions, Exp 1, Exp 5 and Exp 6, have been selected and discussed here. No NH4HCO3 was added in Exp 1, but 6 g and 17.4 g NH4HCO3 were added into reaction solution in Exp 5 and Exp 6, respectively. It has been shown in Figure 6 that the dissolving NH4HCO3 into solution could dramatically increase the system CO2 removal capability. Instead of 20 min in Exp 1, CO2 removal efficiency could be stabilized at 98% for more than 1 hr in Exp 5 and Exp 6. More NH4HCO3 is added into solution, more stable is the CO2 removal efficiency. Conclusions It is an attractive way to remove CO2 by ammonium solution scrubbing. CO2 removal efficiencies are quite stable in the range of 95% to 99% under proper operation conditions. The ammonium solution with higher carbonized degree will have better ability to absorb CO2. Higher CO2 removal efficiency could be lasted for much longer time. Recycle of spent-solution into reactor or adding ammonium salt into the reagent will help to maintain the carbonized degree of solution, and further enhance the ammonium solution CO2 absorption capability. The addition of MEA could stabilize the CO2 removal processes, but no significant improvement on ammonium absorption capability. References: 1. Arrhenius, S. On the Influence of Carbonic Acid in the Air upon the Temperature of the Ground, The London, Edinburgh and Dublin Philosophical Magazine and Journal of Science, Fifth series, 1896, 41 237-277. 2. IPCC Intergovernmental Panel on Climate Change, Climate Change. The IPCC Scientific Assessment, Cambridge University Press, 1990, XXXIX +358. 3. UNFCCC, Kyoto Protocol to the United Nations Framework Convention on Climate Change, FCCC/CP, 1997, L.7/Add.1, Bonn.

4. Global Climate Change Initiative (GCCI) information available on the web at http://www.epa.gov, President Bush Announces Clear Skies & Global Climate Change Initiatives, 2002, Feb. 14, http://www.epa.gov/epahome/headline2_021402.htm 5. Herzog, H.; Eliasson, B.; Kaarstad, O. (2000), Capturing Greenhouse Gases, Scientific American, 2000, 282(2), 72-79. 6. Report of the Conference of the Parties on its Third Session Held at Kyoto, Dec.1-11, 1997, Kyoto Protocol, FCCC/CP/1997/ 7/Add.1, United Nations, New York, NY. 7. Andres L.; Bo L.; Tobias, M. A Fluidized-bed Combustion Process with Inherent CO2 Separation; Application of Chemical-looping Combustion, Chem. Eng. Sci., 2001, 56, 3101-3113. 8. White, C.M.; Strazisar, B.R.; Evan JJ. G.; James S.H.; Henry W.P. Separation and Capture of CO2 from Large Stationary Sources and Sequestration in Geological Formations Coalbeds and Deep Saline Aquifers, J. Air & Waste Manage. Assoc. 2003, 53, 645-715. 9. Rosenberg, N.J.; Izaurralde, R.C. Storing Carbon in Agricultural Soils: a Multi-purpose Environmental Strategy, Kluwer Academic Publishers, 2001. 10. Schnitzer, M. Soil Organic Matter - The Next 75 Years, Soil Sci., 1991, 151, 41-58. 11. Freund, P. Abatement and Mitigation of Carbon Dioxide Emissions from Power Generation, Power-gen 98 Conference, Milan, June, 1998, http://www.ieagreen.org.uk/pge98.htm 12. Yeh, J. T., et al. Aqua Ammonia Process for Simultaneous Reduction of CO2, SO2, and NOx, Proceedings, 19th Annual International Pittsburgh Coal Conference, Pittsburgh, PA, September 23-27, 2002 . 13. Bai, H.; Yeh, A. C. Removal of CO2 Greenhouse Gas by Ammonia Scrubbing, Industrial and Engineering Chemistry Research, 1997, 36(6), 2490-2493. 14. Yeh, A.C. and H.Bai, Comparison of Ammonium and Monoethanolamine Solvents to Reduce CO2 Greenhouse Gas Emission, The Science of Total Environment, 1999, 228, 121-133. 15. Scott M. Smouse, James M. Ekmann, and et al. Experimental study to capture CO2 in the flue gas from a coal-fired research facility by spraying aqueous ammonium to produce a modified NH4HCO3 fertilizer, Second annual conference on carbon sequestration: Developing & Validating the technology base to reduce carbon intensity, May 5-9, 2003, Virginia