Chemical Engineering Science 107 (2014) 218–226
Contents lists available at ScienceDirect
Chemical Engineering Science journal homepage: www.elsevier.com/locate/ces
Amino acid salts for CO2 capture at flue gas temperatures Chiao-Chien Wei n, Graeme Puxty, Paul Feron CSIRO Energy Technology, PO Box 330, Newcastle NSW 2300, Australia
H I G H L I G H T S
To To To To To
develop novel absorbents for CO2 capture at high temperatures. determine the physical properties of the novel absorbents at high temperatures. determine the CO2 solubility of the novel absorbents at high temperatures. investigate the CO2 kinetics of the novel absorbents at high temperatures. model the absorption reaction rate of the novel absorbent at high temperatures.
art ic l e i nf o
a b s t r a c t
Article history: Received 5 March 2013 Received in revised form 13 November 2013 Accepted 20 November 2013 Available online 10 December 2013
The amino acid salt potassium taurate has potential for use as a high-temperature absorbent for postcombustion CO2 capture, because of its low volatility and high absorption rate. In this study, the densities and viscosities of 2–6 M taurate solution were determined over the temperature range of 293–353 K. We found that the CO2 solubility of taurate solutions, measured using a stirred-cell reactor, is comparable to that of alkanolamines at high temperatures. The absorption rate of CO2 into CO2-free and CO2-loaded taurate solutions was determined using a wetted-wall column. The KG of 4 M taurate at 353 K is similar in magnitude to the KG of 7 m monoethanolamine (MEA) at 313 K. We also found that the KG of taurate decreased with increased CO2 loading, although the KG values of taurate solutions are still comparable to CO2-loaded 7 m MEA solution. The reaction rate constant of taurate carbamate formation in this work agrees well with published values. Crown Copyright & 2013 Published by Elsevier Ltd. All rights reserved.
Keywords: CO2 capture Overall mass transfer coefficient Amino acid salts Kinetics CO2 solubility High temperature absorbent
1. Introduction A general consensus from the work of many climate scientists indicates that global warming and climate change are the result of anthropogenic emissions of greenhouse gases. Carbon dioxide (CO2) is thought to be the most important of these, with its major source being the combustion of fossil fuels (coal, natural gas and oil). Fossil fuels currently supply more than 85% of world's energy supply, approximately 40% of which is produced in power plants (Figueroa et al., 2008; Energy Information Administration, 2011). The technology of post-combustion capture (PCC) is well recognised by the government and industry as an effective way of absorbing 80–90% of CO2 emissions from fossil fuel-fired power plants (MacDowell et al., 2010). Captured CO2 can be stored in depleted oil and gas fields, deep saline aquifers and unmineable coal seams, thereby reducing CO2 emissions to the atmosphere.
n
Corresponding author. Tel.: þ 886 3 483 7701×2318. E-mail address:
[email protected] (C.-C. Wei).
Most current PCC processes use liquid absorbents, such as aqueous ammonia or alkanolamine solutions, which typically absorb CO2 at temperatures of 283–313 K (Bandyopadhyay, 2011). The temperatures of flue gases in fossil fuel-fired power plants range from 39 to 433 393 to 433 K are much higher than the temperature of absorption in PCC. Although the temperatures of flue gases can be reduced through the flue-gas desulfurization (FGD) process, there is no FGD in any Australian power station. The additional cooling systems and equipment required to cool flue gases add to the cost, water and energy consumption of PCC. Peeters et al. (2007) calculated that the capital cost and energy penalty for the additional cooling system and pumps required for PCC in a natural gas-fired power station are 3% and 10%, respectively. Likewise, Fisher et al. (2005) found that the cost of the additional cooling system and pumps in a coal-fired power station is about 4% of total equipment cost, and the energy penalty is about 10% of total energy consumption. Absorbents that can capture CO2 as close to flue gas temperature as possible, clearly constitute a technological breakthrough and promise great potential for a more economic and energy-efficient
0009-2509/$ - see front matter Crown Copyright & 2013 Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ces.2013.11.034
C.-C. Wei et al. / Chemical Engineering Science 107 (2014) 218–226
capture process. In addition, the mass transfer between flue gases and absorbents could be faster at high temperatures, due to the lower viscosity of absorbents and faster CO2 absorption kinetics under high-temperature conditions. Most commercial absorbents are amine-based solvents. They are not suitable to absorb CO2 at high temperatures (4333 K), due to their volatility and thermal degradation (Lepaumier et al., 2010). Recently, researchers have found that functionalised ionic liquids (ILs) can maintain their CO2 solubility at high temperatures. Wang et al. (2010) found that a mixture of ILs and superbases can absorb CO2 at temperatures up to 353 K. A new type of polyamine-based IL can absorb CO2 at temperatures of 383–403 K (Ren et al., 2012). However, it takes at least 30 min to reach the maximum CO2 capacity in ILs, which is much slower than in amine solutions. In addition, the presence of water in the gas stream is a critical issue; a dry environment is required for most IL systems to absorb CO2. This would be a challenge in practice, as power plant flue gas is saturated with water. The complex synthesis procedure of ILs also means that they cost much more than amine solutions. These issues mean that in the near term, ILs will not be practically useful as CO2 absorbents. Amino acid salts have been widely studied as absorbents in the field of CO 2 capture, due to their low volatility and resistance to oxygen degradation in the absorption process (Kumar et al., 2003a; Kumar et al., 2003b; van Holst et al., 2009; Vaidya et al., 2010; Simons et al., 2010; Kumar et al., 2003c). van Holst et al. (2009) investigated the apparent rate constants for several amino acid salts at 298 K to find suitable absorbents for CO2 capture. They found that amino acid solutions such as glycinate, prolinate, sarcosinate and taurate exhibit relatively high reaction rate constants that are similar to monoethanolamine (MEA) solutions. Kumar et al. (2003b) measured the solubility of CO 2 in taurate solution at 298 K and 313 K at CO 2 partial pressures ranging from 0.1 to 6 kPa. They also investigated the kinetics of CO 2 absorption in taurate solution at temperatures of 285–305 K using a stirred-cell reactor and determined the reaction rate constant of taurate. Portugal et al. (2007) compared the overall kinetic constant of CO2 absorption in glycinate and MEA solutions, and found the absorption rate in glycinate solution is faster than in MEA solution. They also determined the density and viscosity of glycinate solutions at concentrations of 0.1–3 M and temperatures of 293–313 K. Knuutila et al. (2011) studied the CO2 absorption kinetics of an amino acid salt solution, sarcosinate, in a laboratoryscale pilot plant. They found that although the absorption rate of sarcosinate solution is faster than MEA, sarcosinate requires a higher reflux ratio and desorption temperature than MEA in the reboiler and stripper for CO2 desorption. Siemens Energy tested a commercial absorbent based on a functional amino acid salt solution in an industrial-scale pilot plant in Germany at 298 K and 313 K. Compared with MEA solution, the amino acid salt solution has near-zero fugitive emissions, less corrosion in equipment materials, and very little oxygen degradation (Jockenhoevel et al., 2009; Jockenhoevel and Schneider, 2011). The studies cited above only investigated the properties of amino acid salts at temperatures of 283–333 K; very few studies have used amino acid salts to absorb CO2 at high temperatures. As noted earlier, CO2 absorption at high temperatures could increase reboiler heat duty, but the capital cost of additional cooling systems and the extra space required for these systems also needs to be considered. This study focused on the performance of CO2 capture at temperatures between 323 K and 373 K using potassium taurate solution (2–6 M). We chose taurate solution due to its relatively low cost, low volatility and thermal stability at high temperature. The physical properties of taurate solutions, such as density and viscosity, were also measured.
219
2. Mass transfer model and reaction rate constant determination A software tool implemented in the Matlabs (2011) programming environment describes diffusion and chemical reactions in a falling, thin, liquid film exposed to a gas based on penetration theory. This was achieved by solving the partial differential equations and nonlinear simultaneous equations that define the diffusion, reaction and equilibrium processes occurring in the film as a function of time and film depth (Puxty and Rowland, 2011). Axial dispersion and heating effects are neglected and the equations are solved in one dimension perpendicular to the gas–liquid interface. It is also assumed that physical properties such as viscosity and density remain invariant within the film and as a function of absorbed CO2. The system of partial differential equations to be solved are defined as a combination of Fick's law and chemical reactions (Cussler, 2009; Danckwerts, 1970): ∂c i ∂2 c ¼ Di 2i r i ∂t ∂x
ð1Þ
where Di is the diffusion coefficient of species i (m2 s 1), c i is the concentration of species i in molarity (M), x is the distance from the gas–liquid interface (m), t is time (s) and r i is the rate of formation or destruction of i by chemical reaction (M s 1). The Matlabs function pdepe, which relies on the method of lines, was used to numerically solve the complete system of partial differential equations for a particular chemical system, assuming slab geometry. The pdepe function automatically determines the grid spacing in both time and space to achieve results of a specified accuracy. The reaction rate r i consists of ordinary differential equations whose definition is a function of the chemical reactions that occur in the liquid. Any fast equilibrium, such as protonation equilibria, where equilibration can be considered instantaneous relative to kinetically defined reactions, need to be coupled to these differential equations. This is necessary to provide the complete speciation of the chemical system, which is required to solve the ordinary differential equations. The speciation in a system of equilibria can be solved by finding the roots of nonlinear simultaneous equations (Maeder and Neuhold, 2007). The chemical species and reactions used to derive the reaction rate for each species are given in Table 1. A detailed description of how this derivation is done can be found in (Puxty and Rowland, 2011). Because concentrated taurate solutions are being considered (up to 6 M) in this work, the concentrations of ions can become large at high CO2 loadings. As a consequence, it is necessary to take into account non-ideal behaviour due to changes in activity of charged species. In the calculation of the differential and simultaneous equations, the values of rate and equilibrium constants were corrected for ion activity by calculating activity coefficients (γ i ) with the form of the Debeye–Hückel equation of Eq. (2), and using activities (ai ¼ ci γ i ) rather than concentrations. More complex and accurate expressions for activity that include interaction parameters between ion pairs were not used, because these parameters are unknown for taurate. pffiffi z2i A I pffiffi log 10 γ i ¼ ð2Þ 1 þ 1:5ρ 1=2 I where A ¼ ð1:8248 106 =ðeT Þ3=2 Þ (M) is the Debeye–Hückel law slope (Helgeson et al., 1981), ρ is the density of water (kg dm 3) (Peggs and Bettin, 2008), I is the ionic strength (M) and zi is the charge of species i. The time over which the partial differential Eq. (1) is integrated represents the exposure time te (s) of an element of the liquid to the gas. The distance over which it is integrated represents
220
C.-C. Wei et al. / Chemical Engineering Science 107 (2014) 218–226
Table 1 Chemical equations and associated rate and equilibrium constants used in the film model to calculate the CO2 absorption flux. k3 and k 3 were regressed in this work and k3 is compared to literature values.
CO2 þ H2 O2
k1 ;k 1
HCO3
k1 ¼ e þH
þ
ð22:66 7799Þ T
(M 1 s 1) (Soli and Byrne, ½H2 O 5251:43 36:7816logT þ 102:268 T k1 K1
logK 1 ¼ k1 ¼
k2 ¼ 3:279 1013 e
CO2 þ OH 2k2 ;k 2 HCO3
K2 ¼ K1K6;
CO2 þ AA2
AACOO þ H
þ
12
k3 ¼ 3:23 10 e
6075 T
5700 T
k 3 ¼ 2:935 1018 e H þ CO23 2K 4 HCO3 þ þ K5
logK 3 ¼
H þ þ OH 2K 6 H2 O
logK 5 ¼
þ
H þ AA2
AAH
logK 5 ¼
the film thickness xf (m). Following integration of the system of equations describing the reactions and diffusion, the timeaveraged CO2 flux across the gas–liquid interface (at x ¼ 0) can easily be calculated by Eq. (3) (Samanta and Bandyopadhyay, 2007). N CO2 ¼
DCO2 te
Z
te 0
∂c CO2 dt ∂x
ð3Þ
where NCO2 is the time averaged flux (mol m 2 s 1) and DCO2 is the diffusion coefficient of CO2 (m s-1). The exposure time and the film thickness in a falling, thin film can be calculated by the following two Eqs. (4) and (5). te ¼
2=3 2h 3μ 1=3 πd 3 gρ ϑ
xf ¼
1=3 3μϑ πgdρ
ð4Þ
ð5Þ
where h is the height over which the film falls (cm), μ is the viscosity of the liquid (Poise), g is acceleration due to gravity (981 cm s 2), d is the column diameter (cm), ϑ is the liquid flow rate (cm3 s 1) and ρ is the density of the liquid (g cm 3). The diffusion coefficient of CO2 in the liquid phase was calculated using a correlation of N2O diffusivity as a function of viscosity in aqueous amines (Cullinane and Rochelle, 2006) and the N2O analogy. The N2O analogy states that the ratio of diffusion coefficients between N2O and CO2 in water is equal to that in an aqueous solution of larger viscosity. The diffusion coefficients of taurate in aqueous solution were calculated using the Wilke– Chang correlation (Wilke and Chang, 1955). The film model was solved using the chemical reactions and associated reaction rate and equilibrium constants given in Table 1. The film model was regressed against all the wetted-wall flux data (both CO2 free and CO2 loaded) by optimising the values of the forward and reverse reaction rate constants of taurate carbamate formation, k3 and k-3, respectively. Note that k 3 was only regressed from measurements with a CO2 loading. The regression was carried out at each measurement temperature using the Newton–Gauss–Levenberg/Marquardt method as described in Puxty et al. (2006), according to the objective function of Eq. (6). The temperature dependence of both k3 and k 3 was then fitted to an Arrhenius-type equation. 2 min ðN CO2 N calc CO2 Þ
k3 ;k 3
ð6Þ
2 1 where N calc s ). CO2 is the model-calculated absorption flux (mol m
(Edwards et al., 1978)
(M 1s 1) (Kucka et al., 2003)
k 2 ¼ Kk22
k3 ¼ 2:695 1012 e k3 ;k 3
6613 T
2002)
(M-1s-1) regressed in this work, T ¼ 323–353 K
(Kumar et al., 2003b), T ¼285-305 K 6500 T
(M-1s-1) regressed in this work, T ¼323–353 K
2902:39 0:02379T þ 6:4980 (Harned and Samuel, 1941) T 1:801 þ 2166 T regressed from data in (Hamborg et al., 2007) 5839:48 22:4773logT þ 61:2060 þ logc H2 O (Edwards et al., T
1978)
3. Experimental 3.1. Preparation of taurate solutions Taurine (Z99%) and potassium hydroxide (99.99%) were purchased from Sigma-Aldrich and were used directly without further purification. Unloaded taurate solution was prepared by neutralising taurine with an equimolar quantity of potassium hydroxide in deionised water. The neutralisation reaction was carried out in a water bath to maintain constant cooling. Loaded solution was prepared by bubbling CO2 (99.5%, BOC Gas Australia) from a gas cylinder through a fritted bubbler into a batch of CO2 free solution. The solution was placed in a glass bottle and immersed in a water bath. The water bath was then placed on a balance (GX-6100, A&D Weighing). The amount of CO2 loading was thus determined by CO2 added mass. The top of the glass bottle was connected to a cooling condenser to condense any vapour produced during the exothermic absorption reaction. 3.2. Density and viscosity measurements Densities of unloaded and loaded taurate solutions were measured using a density metre (DMA 38, Anton Paar) at temperatures between 298 K and 313 K with an error of 70.001 g cm 3. Due to the limit of the density metre, the densities of taurate solutions at temperatures above 313 K was determined by a traditional method of measuring the mass and the volume of the solution. This was carried out by adding the taurate solutions to a volumetric flask (Pyrex) and weighing the mass of the solution using a balance (GR 300, A&D Weighing). The volume of the volumetric flask had been calibrated at temperatures from 313 K to 353 K using deionized water. The expanded volume of the flask at different temperatures was calculated using the mass of water in the flask divided by the density of water obtained from Perry's Handbook (1997). Dynamic viscosities of the taurate solutions were measured using a viscometer (AMVn, Anton Paar) at temperatures between 298 K and 353 K with a specified repeatability of o0.1%. The integrated temperature of the measurements was specified with a resolution of 70.01 K and the accuracy was 70.05 K. 3.3. Mass transfer measurements – wetted-wall column (WWC) The WWC consisted of a stainless steel column with an effective length of 8.21 cm and diameter of 1.27 cm. Liquid stored in a reservoir was pumped up the inside of the column and through boreholes at the top. It then flowed down the outer surface of the column, creating a thin liquid film. Once the liquid
C.-C. Wei et al. / Chemical Engineering Science 107 (2014) 218–226
NCO2 ¼ K G ðPCO2 PnCO2 Þ
ð7Þ
where PCO2 (kPa) is the logarithmic mean of the CO2 inlet and outlet partial pressure, and PnCO2 (kPa) is the equilibrium CO2 partial pressure.
Fig. 2. Schematic process showing the wetted-wall column apparatus used to determine the overall mass transfer coefficient (KG) of the absorbent.
30
25
Nco2, mmol s-1 m-2
flow reached the base of the column, it was returned to the reservoir to form a closed loop. The column was located within a glass jacket cover with an internal diameter of 2.2 cm and an external diameter of 4.8 cm. Water was circulated between the glass jacket and a water bath to maintain a constant temperature for the entire system. To generate an even and uniform liquid film on the outer surface of the column, the column was rinsed by circulating the absorbents for 30 min without the glass jacket. We can see the liquid film closely and make sure the liquid film distributed uniformly and homogeneously on the column. After that, we covered the glass jacket and started to conduct the WWC experiments. It is also important to check the liquid film again when the experimental conditions were changed. The gas stream was a mixture of CO2 and N2. The proportion of each gas was controlled by Bronkhorst mass flow controllers to mimic typical CO2 concentrations found in different sections of an absorption column. The mixture was the first passed through a stainless steel coil immersed in a water bath. It was then introduced to a saturator, also immersed in the water bath, which contained a fritted bubbler under 23 cm of water. The gas flow then entered the base of the internal chamber of the glass jacket and moved upwards, contacting the liquid film on the column surface in a countercurrent fashion, and was exhausted from the top of the glass jacket. The contact area between the liquid and the gas was fixed as illustrated in Fig. 1. The liquid flow rate was maintained at 121.4 ml/min. The gas flow rate was maintained at 3 L/min in this study. It should be noted that the gas flow rate has been tested from 1 L/min to 5 L/min. We found the overall mass transfer coefficient showed no significantly difference at the different gas flow rates. This is because the mass transfer coefficient of the gas phase (kg) is much larger than that of the liquid phase (kl) i.e. the mass transfer resistance in the gas phase can be negligible, the overall mass transfer coefficient in the WWC is mainly dominated by kl. A schematic of the wetted-wall column is show in Fig. 2. The amount of CO2 absorbed from the gas phase into the liquid phase was determined by measuring the CO2 content of the gas entering and exiting the column. This information, combined with the surface area of contact between the liquid film and the gas, allowed the determination of the CO2 absorption flux, NCO2 (mmol s 1 m 2). The overall mass transfer coefficient, KG (mmol s 1 m 2 kPa 1), was then determined via the equation below:
221
y = 2.395x + 0.04812
20
15
10
5
0 0
2
4
6
8
10
12
Pco2, kPa Fig. 3. CO2 absorption flux (Nco2) versus applied CO2 partial pressure (Pco2) for 2 M taurate at 353 K.
A plot of NCO2 versus PCO2 yields a straight line with a slope equal to KG. An example is shown for 2 M taurate at 353 K in Fig. 3. The experimental data in Fig. 3 including CO2 absorption flux and the CO2 inlet and outlet partial pressure are also listed in Table 2. The wetted-wall column experiments were very time and chemicals consuming work. To make sure the increasing CO2 loading in absorbent did not affect the experimental results, we always used 500 ml fresh absorbents for each CO2 partial pressure. A more detailed description of the experimental setup and procedure has been reported by Puxty et al. (2010). The KG of 2, 4 and 6 M taurate solution was determined. For each solution, the CO2 loading (mol CO2/mol taurate) was varied over the values of 0, 0.1, 0.2 and 0.3. The WWC experimental data are shown in Appendix A. The equilibrium CO2 partial pressure of the CO2-loaded solutions was determined using the stirred-cell reactor as described in Section 3.4. 3.4. Stirred-cell reactor, VLE
Fig. 1. Diagram of the wetted-wall column covered by a jacketed glass.
A stirred-cell reactor was used to investigate the solubility of CO2 in taurate solution. The apparatus is shown in Fig. 4. The reactor (Parr model 5104) is a closed system containing both gas and liquid phases. The temperature of the reactor was maintained by water circulated between the glass jacket surrounding the reactor content and a water bath. Once loaded with liquid, the
222
C.-C. Wei et al. / Chemical Engineering Science 107 (2014) 218–226
Table 2 WWC experimental data for 2 M taurate at 353 K. CO2 Inlet partial pressure (kPa)
CO2 outlet partial pressure (kPa)
Pco2 (log mean of the inlet CO2 absorption and outlet CO2 partial flux (Nco2, pressure) (kPa) mmol s 1 m 2)
2.460 4.710 6.965 9.296 11.675
1.980 3.895 5.811 7.882 9.808
2.211 4.290 6.371 8.570 10.715
5.83 10.3 15.1 19.4 26.7
five known amounts of CO2 into the empty reactor. Using the measured pressures and temperatures, and known compressibility factor taken from Perry's Handbook, the volume was determined using Eq. (9). A more detailed description of the stirred-cell setup and experimental process has been reported by Puxty et al. (2009). The VLE experimental data are shown in Appendix B.
4. Results and discussion 4.1. Density and viscosity Fig. 5 and Table 3 show the densities of 30% MEA solution and taurate solutions with concentrations from 2 to 6 M at temperatures of 353–323 K. The densities of the taurate solutions range from 1.1 to 1.4 g cm 3, while the densities of 30% MEA range from 0.98 to 1.01 g cm 3. The densities of the taurate solutions increased with increasing taurate concentrations, and decreased with increasing temperatures due to the volume expansion of the solution. The viscosities of taurate solutions were also determined in this study. An absorbent with lower viscosity can result in less interfacial mass-transfer resistance between gas and liquid phases. This means that the diffusion coefficient of CO2 in the liquid phase can be increased, thereby increasing the CO2 absorption rate. Fig. 6 plots viscosities (η) of 30% MEA and taurate solutions on a log scale against 1/T. The experimental data of the viscosities are shown in Table 4. The effect of temperature on viscosities of MEA and taurate solutions can be described by the Arrhenius relationship (Eq. (10)): η ¼ Ae B=RT
Fig. 4. Schematic process showing the stirred-cell apparatus used to determine CO2 loading as a function of pressure.
reactor was purged with nitrogen at atmospheric pressure before starting the experiment. CO2 was injected from a CO2 reservoir which suspended on a balance into the reactor. It should be noted that the weighing balance has the accuracy of 70.01 g. The accuracy of the CO2 amount used for the calculation of CO2 solubility was 70.1 g which was in the accuracy range of the balance. Following injection, the pressure in the reactor increased and then decreased as CO2 was absorbed into the liquid phase. The variation of pressure was monitored until the pressure reached an equilibrium value. The liquid phase was stirred by a Teflon-coated, stainless-steel stirrer. The amount of CO2 in the liquid phase was determined by the equation below: G L nTol CO2 ¼ nCO2 þ nCO2
where η is the viscosity (mPa s), A (mPa s) and B (kJ/mole) are constants for the given liquid, and R is the gas constant (8.3144 J/mol K). Table 5 shows the constants for the Arrhenius relationship at different temperatures and the correlation coefficients (r2) of the plots in Fig. 6. The viscosities of 2 M taurate solutions were lower than 30% MEA, and the viscosities of 4 M taurate solutions were similar to 30% MEA at temperatures of 293–353 K. This indicates that the diffusion coefficient of CO2 in the taurate solutions will be larger than or similar to 30% MEA solution. Thus, the liquid-side masstransfer resistance in taurate solutions at higher temperatures could be smaller than in 30% MEA solution at 313 K (Wu et al., 2011). The viscosities of taurate solutions remained constant after loading with CO2, whereas the viscosity of CO2-loaded 30% MEA solution increased. It is believed that the increase of viscosity in 1.4
ð8Þ
nTol CO2 is
Z¼
PV m RT
ð9Þ
where the P is equilibrium pressure, R is the gas constant and Vm is the molar volume in the gas phase. Vm can be obtained by subtracting the liquid volume from the total reactor volume, which includes the reactor, inserts, valves and tubing. To determine the total reactor volume, a calibration was carried out by injecting
1.3
Density (g cm-3)
the total moles of CO2 injected into the reactor, where the nGCO2 is the moles of CO2 in the gas phase and nLCO2 is the moles of CO2 in the liquid phase. nTol CO2 was determined by measuring the mass loss of the CO2 reservoir. nGCO2 was determined by the compressibility factor (Z) below (Perry's, 1997), the CO2 partial pressure (determined from the difference between the initial and final pressure), and the gas volume as per the equation below:
ð10Þ
6M Taurate 4M Taurate 2M Taurate 30% MEA
1.2
1.1
1.0
0.9 290
300
310
320
330
340
350
360
Temperature (K) Fig. 5. Densities of taurate solutions and 30% MEA at temperatures of 298–353 K.
C.-C. Wei et al. / Chemical Engineering Science 107 (2014) 218–226
Table 3 Density ρ (g cm 3) for 2 M, 4 M and 6 M taurate solutions and 30% (wt/wt) MEA at temperatures of 298–353 K. ρ (g cm 3)
T (K)
298 313 323 333 343 353
2 M taurate
4 M taurate
6 M taurate
30% (w/w) MEA
– – 1.124 1.120 1.114 1.109
– – 1.255 1.249 1.244 1.238
– – 1.371 1.367 1.362 1.365
1.011 1.003 0.998 – 0.986 0.979
223
the CO2-loaded 30% MEA solution resulted from the increasing ionic concentration and the increasing strength of electrostatic interaction. In amino acid salt solutions, the interaction between CO2 and solution is dominated by hydrogen bonds (Jeffrey and Maluszynska, 1982). Due to the complex crystal structures of amino acid salts, the hydrogen bonds are short, which stabilises the viscosity of the solutions. The characteristic of stable viscosity is an advantage of taurate solutions for CO2 capture, because the diffusion coefficient of CO2 in the solutions remains constant as CO2 is absorbed.
4.2. CO2 solubility
4 M taurate
6 M taurate
CO2 solubility in taurate solution was investigated using a stirred-cell reactor. Fig. 7 shows the CO2 solubility in taurate solutions at 333 K. At a given CO2/amine molar ratio, the CO2 partial pressure increased with decreasing taurate concentration. This is because less free taurate molecules were available to react with CO2 in lower concentrations of taurate solution, resulting in a higher CO2 partial pressure in the gas phase i.e. lower CO2 solubility. Fig. 8 shows CO2 solubility in 4 M and 6 M taurate solution at temperatures of 333–373 K. The CO2 solubility in the taurate solutions decreased with increasing temperature. A similar trend can be also found in another amino acid salt: glycinate solution. Portugal et al. (2009) reported that the CO2 solubility in 1 M potassium glycinate significantly reduced when absorption temperatures increased from 293 K to 323 K. However, in our study, the CO2 solubility increased with increasing taurate concentration, indicating that a high concentration of taurate solution could be used for high-temperature absorption. The correlation between CO2 solubility and absorbent concentration in taurate solution is different from that in MEA solution. Lee et al. (1976) found that CO2 solubility reduced with increasing MEA concentration. As shown in Fig. 9, CO2 solubility in MEA solution decreased as MEA concentration increased at 373 K. The decrease in CO2 solubility in MEA solution at high temperature can result from evaporation of MEA. The results of our study demonstrate superior CO2 solubility in taurate solution at high temperature compared with MEA solution.
4.3. Overall mass transfer coefficient (KG)
Fig. 6. Viscosities of taurate solutions and 30% MEA at temperatures of 291–353 K.
Table 4 Viscosity η (cp) for CO2 loaded MEA and taurate solutions at temperatures of 293 K353 K. η (cp)
T (K) 30% MEA
2 M taurate
CO2 loading, α (nco2/namine)
293 298 303 313 323 333 343 353
0
0.2
0.4
0
0.1
0.3
0
0
– 2.48 – 1.67 1.33 – 0.92 0.77
– 2.9 – 2.0 1.6 – 1.1 0.9
– 3.5 – 2.4 1.9 – 1.3 1.1
1.465 – 1.235 1.080 0.971 0.892 0.815 0.766
1.530 – 1.286 1.121 1.006 0.921 0.848 0.787
1.591 – 1.333 1.160 1.038 0.949 0.882 0.811
3.002 – 2.366 1.935 1.635 1.421 1.265 1.147
10.520 – 7.491 5.587 4.328 3.459 2.842 2.390
Fig. 10 shows the overall mass transfer coefficient (KG) of taurate solution (2–6 M) at temperatures of 323–353 K. KG is plotted against the equilibrium CO2 partial pressure as determined from the CO2 solubility data. As a comparison, the KG of 7 m MEA
Table 5 Constants of Arrhenius relationship for MEA and taurate solutions and the correlation coefficients of Eq. (10). Absorbent
A (mPa∙s)
B (kJ/mole)
R2
30% 30% 30% 2M 2M 2M 4M 6M
0.00143 0.00170 0.00207 0.03247 0.03153 0.03178 0.01011 0.00168
18.42 18.41 18.38 9.19 9.37 9.44 13.76 21.19
0.998 0.999 0.999 0.989 0.989 0.988 0.990 0.996
MEA MEAþ 0.2 CO2 MEAþ 0.4 CO2 taurate taurateþ0.1 CO2 taurateþ0.3 CO2 taurate taurate
Fig. 7. Solubility of CO2 in taurate solutions at 333 K.
224
C.-C. Wei et al. / Chemical Engineering Science 107 (2014) 218–226
The value of KG increases with increasing taurate concentration, because more unreacted taurate ions are available to react with CO2. Also, as would be expected, KG values decreased with increasing CO2 loading. As we had determined that the viscosity of taurate solutions was unaffected by CO2 content, the decrease in KG must be primarily due to the reduced availability of unreacted taurate. The decreases of KG were more significant in 6 M taurate solution than in the 2 M and 4 M solutions at the same temperature. In more concentrated taurate solutions, more protonated taurate can be produced, and the unstable protonated form can be precipitated in the solution. Although precipitation was not visually observed during these experiments, the KG of the more concentrated taurate solution can be affected by the unstable protonated form. Fig. 8. Solubility of CO2 in 4 M and 6 M taurate solutions at 333, 353 and 373 K.
10000 1000
Pco2, kPa
100 10 2M Taurate 4M Taurate 6M Taurate 1M MEA (Lee et al., 1976) 5M MEA (Lee et al., 1976)
1 0.1 0.01 0.0
0.2
0.4
0.6
0.8
1.0
CO2/amine, mol/mol Fig. 9. Solubility of CO2 in taurate and MEA solutions at 373 K.
4.3.2. Effect of absorption temperature on overall mass transfer coefficient KG Fig. 11 plots the overall mass transfer coefficient (KG) of taurate solution (2–6 M) at temperatures of 323–353 K against the molar ratio of CO2 over amine groups. The KG values in 2 M and 4 M taurate solutions were measured at absorption temperatures from 323 to 353 K. As protonated taurate can be easily precipitated in solutions with high concentrations of taurate at low temperature, the KG values in 6 M taurate solution was measured from 333 to 353 K. The KG values of the taurate solutions increased with increasing absorption temperature. The KG values in the CO2-free 2–6 M taurate solution at temperatures above 323 K are similar in magnitude to the KG of 7 m MEA at 313 K, which is about 2.8 mmol s 1 m 2 kPa 1 (Puxty et al., 2010). Although the KG values decreased with increased CO2 loading, the KG values of taurate solutions at temperatures above 323 K are still comparable to CO2-loaded 7 m MEA solution at 313 K. The taurate solutions show superior CO2 absorption kinetics at higher temperatures compared with MEA solution. 4.4. Reaction rate constant (k3)
5 6M, 353K 6M, 343K 6M, 333K 6M, 323K 4M, 353K 4M, 343K 4M, 333K 4M, 323K 2M, 353K 2M, 343K 2M, 333K 2M, 323K 7m MEA, 333K (Dugas et al., 2011) 7m MEA, 313K (Dugas et al., 2011)
KG, mmol s-1 m-2 kPa-1
4
3
2
1
The reaction rate constant of taurate carbamate formation at temperatures of 323–353 K are shown in Fig. 12 and are given in Table 1. These were determined by fitting model calculated flux to measured flux by optimising the values k3 and k 3 (the forward and reverse rate constants of carbamate formation) using the film model as described in Section 2. As a comparison, the reaction rate constants in taurate solution at 285.15, 395.15 and 305.15 K determined by Kumar et al. (2003c) are also shown. The agreement between k3 as determined in this work and that of Kumar et al. (2003c) is reasonable, considering this work was carried out 4.0
5
10
15
20
25
30
Pco2, kPa Fig. 10. Overall mass transfer coefficient (KG) against CO2 partial pressure (PCO2) in taurate solutions at temperatures of 323–353 K and in 7 m MEA solution at 313 and 333 K.
( 5 M MEA) solution at 313 and 333 K determined by Dugas and Rochelle (2011) is also shown in Fig. 10.
KG, mmol s-1 m-2 kPa-1
0 0
6M, 353K 6M, 343K 6M, 333K 6M, 323K 4M, 353K 4M, 343K 4M, 333K 4M, 323K 2M, 353K 2M, 343K 2M, 333K 2M, 323K 7m MEA, 333K (Dugas et al., 2011) 7m MEA, 313K (Dugas et al., 2011)
3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 0.0
4.3.1. Effect of taurate concentration on overall mass transfer coefficient KG Fig. 10 shows that the overall mass transfer coefficient (KG) in taurate solutions has been affected by the concentration of taurate.
0.2
0.4
0.6
0.8
1.0
CO2/amine, mol/mol Fig. 11. Overall mass transfer coefficient (KG) against the molar ratio of CO2/amine group in taurate solutions at temperatures of 323–353 K and in 7 m MEA solution at 313 and 333 K.
C.-C. Wei et al. / Chemical Engineering Science 107 (2014) 218–226
225
Table A1 WWC experimental data for taurate solutions. Taurate concentration (M)
T, K
2
323
0 0.1 0.2 0.3
1.916 1.516 1.182 0.764
333
0 0.1 0.2 0.3
2.039 1.765 1.33 0.884
343
0 0.1 0.2 0.3
2.261 1.951 1.412 0.713
353
0 0.1
2.395 2.22
323
0 0.1 0.2
2.087 1.702 1.433
333
0 0.1 0.2
2.27 1.887 1.648
343
0 0.1 0.2
2.605 2.126 1.822
353
0 0.1
3.245 2.412
323 333
0 0 0.1
2.501 3.008 1.798
343
0 0.1
3.165 2.208
353
0 0.1
3.746 2.662
Fig. 12. The forward reaction rate constant of taurate carbamate formation at temperatures of 323–353 K. 4
25 0 loading (measured) 0 loading (calculated) 0.1 loading (measured) 0.1 loading (calculated) 0.3 loading (measured) 0.3 loading (calculated)
Nco2, mmol s-1 m-2
20
15
10 6
5
0 0
2
4
6
8
10
12
14
16
18
Pco2, kPa
CO2 loading (mol co2/mol
amine)
KG (mmol s 1 m 2 kPa 1)
Fig. 13. Measured (points) and model calculated (lines) CO2 absorption flux values at 333 K for 2 M taurate.
from 323 to 353 K, whereas Kumar et al. (2003c) covered 285–305 K. Some differences would be expected due to the use of activities in this work compared to concentrations in Kumar et al. (2003c) in addition to the uncertainty associated with extrapolation from one temperature range to another. In addition, the reaction activation energy in this study is 50.5 KJ mol 1, which is very close to the activation energy reported by Kumar et al. (2003c). This result indicates that the reaction rate of taurate carbamate formation can be increased by increasing absorption temperature. The fast CO2 absorption rate of taurate solution at high temperatures is a benefit of hightemperature absorption. The agreement between measured and model calculated flux values was excellent. Fig. 13 shows measured and calculated data for 2 M taurate at 333 K and CO2 loading of 0, 0.1 and 0.3. This is typical of the agreement across the entire dataset. Some curvature of the relationship between Nco2 and Pco2 occurs particularly for the measurements at high CO2 loading. This behaviour is capture by the model used which is based on penetration theory. It represents the situation where taurate becomes depleted at the gas–liquid interface and diffusion of taurate to the interface limits the flux.
taurate solution is higher than 30% MEA solution, and that the viscosities of 4 M taurate solutions are similar to 30% MEA solution at temperatures of 293–353 K. The CO2 solubility of taurate solutions can be increased by increasing the concentration of taurate, and is comparable to the solubility of alkanolamines at high temperatures. KG increased with an increase of taurate concentration and absorption temperature. The KG values of CO2-free taurate solutions at temperatures above 323 K are similar in magnitude to the KG of 7 m MEA at 313 K. The KG values of CO2-loaded taurate solutions at temperatures of 323–353 K are higher than those of CO2-loaded 7 m MEA at 313 and 323 K. The reaction rate constant of taurate carbamate formation in this work is consistent with published results. In summary, our study has demonstrated that the CO2 solubility and CO2 absorption kinetics of taurate solutions at temperatures up to 353 K are superior to those of MEA solution.
Appendix A See Table A1.
5. Conclusion Appendix B We investigated the physical and kinetic properties of taurate solutions for its application in PCC. We found that the density of
See Table A2.
226
C.-C. Wei et al. / Chemical Engineering Science 107 (2014) 218–226
Table A2 VLE experimental data for taurate solutions. Taurate concentration (M)
T (K)
2
333
0.436 0.821 0.961 1.098
16.1 56.9 82.5 96.8
353
0.361 0.491 0.554 0.606
31.3 189.9 398.6 574.2
373
0.153 0.281 0.362 0.435
17.3 96.9 257.7 514.8
333
0.096 0.246 0.542 0.686
4.7 5.4 11.5 25.8
353
0.056 0.172 0.447 0.603
5.3 6.3 9.5 17.2
373
0.159 0.328 0.496 0.654 0.798 0.897
7 10.7 21.2 51.2 133.6 271.1
333
0.161 0.378 0.599
1.9 2.2 3.2
353
0.236 0.356 0.476 0.731 0.868 0.986
2.2 3.6 4.3 7.3 9.2 14.5
373
0.306 0.508 0.618 0.722 0.823
3.7 6.5 7.6 9.4 10
4
6
CO2 loading (mol co2/mol
Pco2 (kPa) amine)
References Perry's, 1997. Chemical Engineers' Handbook, 7th ed. McGraw-Hill. Matlabs 2011. 2009b ed. MathWorks™, Natick. Bandyopadhyay, A., 2011. Amine versus ammonia absorption of CO2 as a measure of reducing GHG emission: a critical analysis. Clean Technol. Environ. Policy 13, 269–294. Cullinane, J.T., Rochelle, G.T., 2006. Kinetics of carbon dioxide absorption into aqueous potassium carbonate and piperazine. Ind. Eng. Chem. Res. 45, 2531–2545. Cussler, E.L., 2009. Diffusion - Mass Transfer in Fluid Systems, 3rd ed. Cambridge University Press, Cambridge. Danckwerts, P.V., 1970. Gas-Liquid Reactions. McGraw-Hill, New York. Dugas, R.E., Rochelle, G.T., 2011. CO2 absorption rate into concentrated aqueous monoethanolamine and piperazine. J. Chem. Eng. Data 56, 2187–2195. Edwards, T.J., Maurer, G., Newman, J., Prausnitz, J.M., 1978. Vapour-liquid equilibria in multicomponent aqueous solutions of volatile weak electrolytes. Aiche J. 24, 966–976. Energy Information Administration, 2011 International Energy Outlook 2011. DOE/ EIA-0484(2011). Figueroa, J.D., Fout, T., Plasynski, S., McIlvried, H., Srivastava, R.D., 2008. Advances in CO2 capture technology-the U.S. department of energy's carbon sequestration program. Int. J. Greenhouse Gas Control 2, 9–20. Hamborg, E.S., Niederer, J.P.M., Versteeg, G.F., 2007. Dissociation constants and thermodynamic properties of amino acids used in CO2 absorption from (293 to 353) K. J. Chem. Eng. Data 52, 2491–2502. Harned, H.S., Samuel, R.S., 1941. The Ionisation Constant of HCO3- from 0 to 501. J. Am. Chem. Soc. 63, 1706–1709.
Helgeson, H.C., Kirkham, D.H., Flowers, G.C., 1981. Theoretical prediction of the thermodynamic behavior of aqueous electrolytes by high pressures and temperatures; IV, calculation of activity coefficients, osmotic coefficients, and apparent molal and standard and relative partial molal properties to 600 ºC and 5 kb. J. Am. Chem. Soc. 281, 1249–1516. Jeffrey, G.A., Maluszynska, H., 1982. A survey of hydrogen bond geometries in the crystal structures of amino acids. Int. J. Biol. Macromol. 4, 173–185. Jockenhoevel, T., Schneider, R., 2011. Towards commercial application of a secondgeneration post-combustion capture technology - pilot plant validation of the siemens capture process and implementation of a first demonstration case. Energy Procedia 4, 1451–1458. Jockenhoevel, T., Schneider, R., Rode, H., 2009. Development of an economic postcombustion carbon capture process. Energy Procedia 1, 1043–1050. Fisher, K.S., Beitler, C., Rueter, C., Searcy, K., Rochelle, G., Jassim, M., 2005. Integrating MEA Regeneration with CO2 Compression and Peaking to Reduce CO2 Capture Costs. In: Fisher, K.S. (Ed.), US Department of Energy. National Energy Technology Labatory, Pittsburg, Pennsylvania. Knuutila, H., Aronu, U.E., Kvamsdal, H.M., and Chikukwa, A., 2011 Post Combustion CO2 Capture with an Amino Acid Salt. 4 ed., pp. 1550–1557. Kucka, L., Muller, I., Kenig, E.Y., Gorak, A., 2003. On the Modelling and simulation of sour gas absorption by aqueous amine solutions. Chem. Eng. Sci. 58, 3571–3578. Kumar, P.S., Hogendoorn, J.A., Feron, P.H.M., Versteeg, G.F., 2003a. Equilibrium solubility of CO2 in aqueous potassium taurate solutions: part 1 crystallization in carbon dioxide loaded aqueous salt solutions of amino acids. Ind. Eng. Chem. Res. 42, 2832–2840. Kumar, P.S., Hogendoorn, J.A., Timmer, S.J., Feron, P.H.M., Versteeg, G.F., 2003b. Equilibrium solubility of CO2 in aqueous potassium taurate solutions: part 2 experimental VLE data and model. Ind. Eng. Chem. Res. 42, 2841–2852. Kumar, P.S., Hogendoorn, J.A., Versteeg, G.F., Feron, P.H.M., 2003c. Kinetics of the reaction of CO2 with aqueous potassium salt of taurine and glycine. Aiche J. 49, 203–213. Lee, J.I., Otto, F.D., Mather, A.E., 1976. Equilibrium between carbon dioxide and aqueous monoethanolamine solutions. J. Chem. Technol. Biotechnol. 26, 541–549. Lepaumier, H., Martin, S., Picq, D., Delfort, B., Carrette, P.L., 2010. New amines for CO2 capture. III. Effect of alkyl chain length between amine functions on polyamines degradation. Ind. Eng. Chem. Res. 49, 4553–4560. MacDowell, N., Florin, N., Buchard, A., Hallett, J., Galindo, A., Jackson, G., Adjiman, C. S., Williams, C.K., Shah, N., Fennell, P., 2010. An overview of CO2 capture technologies. Energy Environ. Sci. 3, 1645–1669. Maeder, M., Neuhold, Y.M., 2007. Practical Data Analysis in Chemistry. Elsevier, Amsterdam. Peeters, A.N.M., Faaij, A.P.C., Turkenburg, W.C., 2007. Techno-Economic analysis of natural gas combined cycles with post-combustion CO2 absorption, including a detailed evaluation of the development potential. Int. J. Greenhouse Gas Control 1, 396–417. Peggs, S.L., Bettin, H., 2008. Tables of Physical & Chemical Constants. 2.2.1 Densities. Kaye & Laby Online. Version 1.1. National Physical Laboratory. Portugal, A.F., Derks, P.W.J., Versteeg, G.E., Magalhaes, F.D., Mendes, A., 2007. Characterization of potassium glycinate for carbon dioxide absorption purposes. Chem. Eng. Sci. 62, 6534–6547. Portugal, A.F., Sousa, J.M., Magalhaes, F.D., Mendes, A., 2009. Solubility of carbon dioxide in aqueous solutions of amino acid salts. Chem. Eng. Sci. 64, 1993–2002. Puxty, G., Maeder, M., Hungerbuhler, K., 2006. Tutorial on the fitting of kinetics models to multivariate spectroscopic measurements with non-linear leastsquares regression. Chemometr. Intell. Laboratory Systems 81, 149–164. Puxty, G., Rowland, R., 2011. Modeling CO2 mass transfer in amine mixtures: PZ-AMP and PZ-MDEA. Environ. Sci. Technol. 45, 2398–2405. Puxty, G., Rowland, R., Attalla, M., 2010. Comparison of the rate of CO2 absorption into aqueous ammonia and monoethanolamine. Chem. Eng. Sci. 65, 915–922. Puxty, G., Allport, A., Attalla, M., 2009. Vapour liquid equilibria data for a range of new carbon dioxide absorbents. Energy Procedia 1, 941–947. Ren, S., Hou, Y., Wu, W., Tian, S., Liu, W., 2012. CO2 capture from flue gas at high temperatures by new ionic liquids with high capacity. RSC Adv. 2, 2504–2507. Samanta, A., Bandyopadhyay, S.S., 2007. Kinetics and modeling of carbon dioxide absorption into aqueous solutions of piperazine. Chem. Eng. Sci. 62, 7312–7319. Simons, K., Brilman, W., Mengers, H., Nijmeijer, K., Wessling, M., 2010. Kinetics of CO2 absorption in aqueous sarcosine salt solutions: influence of concentration, temperature, and CO2 loading. Ind. Eng. Chem. Res. 49, 9693–9702. Soli, A.L., Byrne, R.H., 2002. CO2 system hydration and dehydration kinetics and the equilibrium CO2/H2CO3 ratio in aqueous NaCl solution. Mar. Chem. 78, 65–73. Vaidya, P.D., Konduru, P., Vaidyanathan, M., Kenig, E.Y., 2010. Kinetics of carbon dioxide removal by aqueous alkaline amino acid salts. Ind. Eng. Chem. Res. 49, 11067–11072. van Holst, J., Versteeg, G.F., Brilman, D.W.F., Hogendoorn, J.A., 2009. Kinetic study of CO2 with various amino acid salts in aqueous solution. Chem. Eng. Sci. 64, 59–68. Wang, C., Luo, H., Luo, X., Li, H., Dai, S., 2010. Equimolar CO 2 capture by imidazolium-based ionic liquids and superbase systems. Green Chem. 12, 2019–2023. Wilke, C.R., Chang, P., 1955. Correlation of diffusion coefficients in dilute solutions. Aiche J. 1, 264–270. Wu, H., Shah, J.K., Tenney, C.M., Rosch, T.W., Maginn, E.J., 2011. Structure and dynamics of neat and CO2-reacted ionic liquid tetrabutylphosphonium 2-Cyanopyrrolide. Ind. Eng. Chem. Res. 50, 8983–8993.
