Vapor-Liquid Equilibrium of Carbon Dioxide in Aqueous Mixtures of

Aug 1, 1994 - Behavior of Reboiler Heat Duty for CO2 Capture Plants Using Regenerable Single and Blended Alkanolamines. Roongrat Sakwattanapong ...
15 downloads 0 Views 409KB Size
Ind. Eng. Chem. Res. 1994,33, 2002-2005

2002

Vapor-Liquid Equilibrium of Carbon Dioxide in Aqueous Mixtures of Monoethanolamine and Methyldiethanolamine Fang-Yuan J o u , Frederick D. Otto, and A l a n

E. Mather'

Department of Chemical Engineering, University of Alberta, Edmonton, Alberta, Canada T6G 2G6

Data for the distribution of carbon dioxide between the vapor and aqueous solutions of four mixtures of monoethanolamine (MEA) and methyldiethanolamine (MDEA) have been obtained at 25,40, 80 and 120 "C over a range of pressures from 100 kPa to 20 MPa. Partial pressures of C02 ranged from 0.001 t o 19 930 kPa. Enthalpies of reaction of COZin the solutions have been calculated from the solubility data. Introduction Aqueous solutions of alkanolamines are used to separate carbon dioxide and hydrogen sulfide from gas streams. The acid gases are absorbed into the solution a t lower temperatures and are desorbed from the solution by heating to higher temperatures. In this way a continuous process for the removal of the acid gases from gas streams results. Monoethanolamine (MEA) and diethanolamine (DEA) have been the most widely used alkanolamines in gas processing. Recently, the tertiary amine methyldiethanolamine (MDEA) has found increased use. It has the advantage that no carbamate is formed and the enthalpy of reaction is smaller than that of either MEA or DEA. Hence the cost of regeneration is lower. However, MDEA solutions react slowlywith C02, and if bulk removal of carbon dioxide is desired, the result is a larger number of trays or an increased height of packing compared with MEA or DEA. Chakravarty et al. (1985) proposed the use of a mixture of MEA and MDEA in aqueous solution to combine the desirable features of both amine solvents. Vickery et al. (1988) and Campbell and Weiland (1989) discussed the use of amine blends (mixtures of MEA and MDEA in aqueous solution). Few data exist for such amine blends. Austgen et al. (1991) presented a model for mixed amine equilibria and data for the solability of CO2 in aqueous mixtures of MDEA with MEA and DEA. Glasscock et al. (1991) presented a comprehensive review of the rate and equilibrium behavior of mixed amines, together with experimental data. Shen and Li (1992) and Li and Shen (1992) reported solubility data for C02 in four mixtures of MDEA and MEA, the total amine concentration being 30 wt 7% in all cases. Zhang et al. (1993) described the modeling of the acid gas removal process and compared their results with plant data. This work was undertaken to resolve some of the questions about the solubility data used for model development and provide data at low concentrations of MEA in the MEA/MDEA blend where the effect of the primary amine is most pronounced. Experimental Section The equipment and procedures used in the experiments are similar to those used in this laboratory in the past (Jou et al., 1982, 1994) and will be described briefly. The equilibrium cell is a Jerguson gauge. It is mounted in an air bath, together with a magnetic pump similar to that devised by Ruska et al. (1970). The pump serves to recirculate the vapor phase and bubble it through the liquid

* To whom correspondence should be addressed.

phase at a rate of up to 100 mL/min. The temperature of the contents of the cell is measured to within 10.1 "C with a calibrated iron-constantan thermocouple, and the pressure in the cell is measured by digital Heise gauges. The error in the gauges was 0.1 % by dead-weight calibration and 0.2 7% by comparison with the measured critical points of propane, carbon dioxide, and hydrogen sulfide (Braker and Mossman, 1980). Monoethanolamine (99+ 7%) and methyldiethanolamine (99%) were obtained from the Aldrich Chemical Co. Carbon dioxide and nitrogen with a minimum purity of 99.8% and 99.998%, respectively, were supplied by Linde. The solutions were prepared by weight using distilled water. About 100 cm3of solution was allowed to enter the evacuated cell a t room temperature. Carbon dioxide was added in amounts monitored by the pressure in the cell. If necessary, nitrogen was added to maintain the total pressure well above atmospheric pressure. When the vapor phase contained nitrogen, it was analyzed to determine the N&02 ratio using a chromatograph with a 3-m X 3.175-mm Porapak QS column operated at 70 "C or Porapak S operated a t 100 "C. The response factor was 1.17 for N2 and 1.00 for COz. In the presence of the amine solution, the overall uncertainty of the vapor pressure analysis is 3% for C02 partial pressures above 10Pa; below that it increases linearly to 30% at 1 Pa. Most of the analyses of the liquid phase were performed using a precipitation of the C02 as BaC03 and subsequent titration with standard 0.1 N HC1 solution. Dilution with 1or 2 M NaOH solution was required for a > 0.5 or pco2 >lo00 kPa. Another analysis was performed using a chromatograph with a 1.63-m X 3.175-mm Chromosorb 104 column and 22 mL/min helium flow. The thermal conductivity detector was set a t 250 "C. A 5-pL sample was injected into the gas chromatograph a t 300 "C. The actual amount of injection was obtained by weighing. Dilution of the sample with 30 w t % diglycolamine (DGA) was required for a > 0.5 andpco, >lo00 kPa. The evolved C02 peak was compared with that of 100 pL of C02 a t barometric pressure and room temperature. The amines were then eluted at 250 "C. The retention time is 0.44 rnin for C02, 1.04 rnin for H20, 3.5 rnin for 100 pL CO2 standard, 8.2 min for MEA, 12.6 min for DGA, and 17.2 min for MDEA. The number of moles versus area counts was used to construct a calibration c w e . The linearity was established for CO2 and HzO > 1 X 1o-S mol with molar response factors of 1.00 for C02 and 1.71 for H20, which do not change with detector temperatures between 150 and 300 "C and helium flow rates from 16 to 32 minl min. The a value was then calculated from the standard

0888-5885/94/2633-2002$04.50/00 1994 American Chemical Society

Ind. Eng. Chem. Res., Vol. 33, No. 8,1994 2003 of 100 pL of COZ,the calibration curve, and the molar response factor multiplied by the molar ratio of H201 amine. The area ratio of H20:MEAMDEA from the sample also served to monitor the amine concentration when compared with that of a designated mixture. The variation of this ratio was kept under 2% so as to minimize error caused by change in MEA concentration, which has a profound effect on the COZpartial pressure. At 120 and 150 "C a slight makeup of HzO and MEA was needed due to vaporization. This was done through a spindle press of 20-mL capacity. The solution was studied for up to 5 consecutive days at 100 and 120 "C or 3 days at 150 O C . The chromatograph of the liquid sample did not show any other peak, indicating a negligible degradation of the amines. A self-consistency test of the COZmeasurement for the liquid phase was performed with a weighed amount of COz in a known amount of 30 wt % DGA at a < 0.3. At room temperature this a < 0.3 corresponds to more than 99.99 % retention of COZin liquid. From three analyses the error was found to be 3% for the BaC03 method and 2% for the gas chromatograph method. The details of the analysis have been presented in the recent paper of Jou et al. (1994).

1 o2

B 10'

a

1 O0

lo-' 0.0

0.2

a

co*

0.4

0.6

0.8

1.0

1.2

(mol of C02/mol of amine)

Figure 1. 1. Comparison of solubility of COZin 30 wt % MDEA solution. Open points and dotted lines: Shen and Li (1992); solid points and solid lies: this work. All lines are empirical.

Results and Discussion In order to calculate the partial pressure of COz in the NdCOz gas mixture, the vapor pressure of the amine must be subtracted from the total pressure, solution (PI") P. PI" was calculated according to Raoult's law from the mole fraction of HzO and amines. The vapor pressure of HzO was taken from the Handbook of Chemistry and Physics (1980), that of MEA was taken from Daubert et al. (1987), and that of MDEA was taken from Daubert and Hutchinson (1990). Data have been obtained for the distribution of COZ between the vapor phase and an aqueous solution containing 30 w t % MDEA at temperatures of 25,40,80, and 120 OC. Partial pressures of COZranged from 0.0025 to 15 000 kPa. The results are presented in Table I (Tables I-VI are supplementary material; see paragraph at end of paper regarding availability of supplementary material) and plotted in Figure 1 for comparison with the data of Shen and Li (1992). The lack of agreement is clearly seen, and the results of Shen and Li do not agree with the earlier work of Jou et al. (1982) or with that of Austgen et al. (1991). Similar disagreement was found with the data for COz in 30 wt % MEA solution (Jou et al., 1994). Possible reasons for the lack of agreement include carrier gas entering the apparatus of Shen and Li. The on-line procedure used by them cannot prevent the carrier gas from entering the system, and depending upon the size of the sampling loop and the number of times injections were made, the contribution of the carrier gas to the total pressure could be significant. Another possible source of error is the slow color change in the indicators used in the titration procedure, phenolphthalein and methyl orange, which is caused by the buffer effect of the amine. This effect is pronounced at high loadings. Data have been obtained for four amine blends containing a total amine concentration of 30 wt %: 1.5% MEA + 28.5% MDEA, 3% MEA + 27% MDEA, 10% MEA + 20 % MDEA, and 20 % MEA + 10 % MDEA. For each solution four temperatures, 25,40, 80, and 120 "C, were considered, and partial pressures of COz ranged from 0.001 to 19 930 kPa. Data for 20% MEA + 10% MDEA at 60, 100, and 150 OC were also obtained. The experi-

1 o3

s

a

10'

1 oo

0.2

0.4

a

coz

0.6

0.8

1.0

1.2

(mol of C02/mol of amine)

Figure 2. Comparison of the solubility of COz in 6 wt % MEA + 24 wt 7% MDEA solution.Points and dotted lines,Li and Shen (1992); solid lines, interpolation from this work.

mental results for the blends are given in Tables 11-V. In order to compare the results with those of Shen and Li and Li and Shen, plots of the data were interpolated to the concentrations used by them. Typical results are shown in Figures 2 and 3. In general, the data of Shen and Li and Li and Shen are in poor agreement with the present work, especially at elevated pressures. The strong effect of the amount of MEA on the partial pressure of COZis shown in Figure 4. At typical absorption conditions, 40 "C, the replacement of 4 wt 7% of the solution by MEA causes a reduction in the partial pressure by a factor of 10. This advantage gradually recedes as the COZloading increases. When LY > 0.7,adding MEA to MDEA solution causes a rise in the partial pressure of COZ,because the MEA carbamate reverts to bicarbonate which is then in equilibrium with COZ.

2004 Ind. Eng. Chem. Res., Vol. 33, No. 8, 1994

I

90

100°C 80°C 60°C1

0 30 w t % M E A 0 20 wt% Y E A 10 w t % MEA

v

80

1 o3

\ h

0"

\ 4

T

10 wL% M D E A

+

20 wl% M D E A

30 wt% MDEA

70

V h

-: *-(

lo2

0

a x

v

2-

0"

a" 10'

60

50

0"

$ I 1 O0

40

30

0.4

0.2

0.6

(mol of COJmol

a

20

1 .o

0.8

I

0.0

0.2

a

of amine)

co2

Figure 3. Comparison of the solubility of COz in 24 w t % MEA +

6 w t % MDEA solution. Pointa and lines, Shen and Li (1992);solid

lines, interpolation from this work.

0.4

co,

0.6

0.8

(mol of COJrnol

1.0

1.2

1.4

of amine)

Figure 5. Enthalpy of solution of COz in mixtures of MEA + MDEA as a function of loading.

amount of MEA in the mixed amine solution. This makes the removal of C02 more difficult from the mixed amine solution compared with that from the MDEA solution. The optimal MEAMDEA ratio depends on the COa residual specification and the stripping energy cost.

1 o2

10'

Conclusions Extensive data for the solubility of carbon dioxide in aqueous solutions of monoethanolamine (MEA) and methyldiethanolamine (MDEA)have been obtained over wide ranges of temperature and pressure. The data serve as a source of information for the modeling of blends of amines for the removal of carbon dioxide from gases.

h

m a 1 oo

x v

0 "

a" lo-'

Acknowledgment

1 o-2

This work was supported financially by the Natural Sciences and Engineering Research Council of Canada under Strategic Grant No. 100777. 0

5

10

15

20

25

30

wt% MEA

Figure 4. Partial pressure of COz as a function of composition at ace, = 0.1.

The enthalpy of solution of carbon dioxide in the amine solutions was obtained by application of the GibbsHelmholtz equation to the solubility data: [d In pld(llT)l, = AHIR A linear relationship was found for all a < 1;above a = 1considerable scatter exists. The overall uncertainty is f2 kJ/mol. In addition to the present data, results for C02 in a 30 wt % MEA solution by Jou et al. (1994)are also plotted in Figure 5. The results are presented in Table VI. In the literature, a value of AH = -62.3kJ/mol at low a in 2.0 and 4.28 kmolIm3 MDEA solutions was reported by Jou et al. (1982)and AH = -85.4 kJ/mol at a = 0.2 was obtained by Lee et al. (1974)in a 30 wt 5% MEA solution. It can be seen from Figure 5 that, at low loading, the enthalpy of the solution of COz is a strong function of the

Nomenclature P = total pressure, kPa p = partial pressure, kPa t = Celsius temperature, "C a = loading, mol of COdmol of amine AH = enthalpy of solution, kJ/mol COz

Supplementary Material Available: Tables I-V containing the experimental data for the vapor-liquid equilibrium and Table VI containing the derived enthalpies of solution (13pages). Ordering information is given on any current masthead page. Literature Cited Austgen, D. M.; Rochelle, G.T.; Chen, C.-C. Model of Vapor-Liquid Equilibria for Aqueous Acid Gas-Ahnolamine Systems. 2. Representation of H& and COz Solubility in Aqueous MDEA and COz Solubility in Aqueous Mixtures of MDEA with MEA or DEA. Ind. Eng.Chem. Res. 1991,30,643-555. Braker, W . , Mossman, A. L., Ede. Matheson Gas Data Book; Matheson Co.: Secaucus, NJ, 1980.

Ind. Eng. Chem.Res., Vol. 33, No. 8, 1994 2005 Campbell, S. W.; Weiland, R. H. Modeling COz Removal By Amine Blends. Presented a t the AIChE 1989 Spring National Meeting, Houston, TX, April 24,1989; paper no. 56d. Chakravarty, T.; Phukan, U. K.; Weiland, R. H. Reaction of Acid Gases with Mixtures of Amines. Chem. Eng. Prog. 1985, 81(4), 32-36. Daubert, T. E.; Hutchison, G. Vapor Pressure of 18 Pure Industrial Chemicals. AIChE Symp. Ser. 1990,86 (279), 93-114. Daubert, T. E.; Jalowka, J. W.; Goren, V. Vapor Pressure of 22 Pure Industrial Chemicals. AIChE Symp. Ser. 1987,83(256), 128-156. Glasscock, D. A.; CritchField, J. E.; Rochelle, G. T. COz Absorption/ Desorption in Mixtures of Methyldiethanolamine with Monoethanolamine or Diethanolamine. Chem. Eng. Sci. 1991, 46, 28292845. Jou, F.-Y.; Mather, A. E.; Otto, F. D. Solubility of Has and COz in Aqueous Methyldiethanolamine Solutions. Ind. Eng. Chem. Process Des. Dev. 1982,21, 539-544. Jou, F.-Y.; Mather, A. E.; Otto, F. D. The Solubility of COz in a 30 Mass Percent Monoethanolamine Solution. Can. J. Chem. Eng. 1994, in press. Lee, J. I.; Otto, F. D.; Mather, A. E. The Solubility of HzS and COz in Aqueous Monoethanolamine Solutions. J. Chem. Eng. Data 1974,52,803-805.

Li, M.-H.; Shen, K.-P. Densities and Solubilities of Solutions of Carbon Dioxide in Water + Monoethanolamine N-Methyldiethanolamine. J. Chem. Eng. Data 1992,37,28&290. Ruska, W. E. A.; Hurt, L. J.; Kobayashi, R. Circulating Pump for High Pressure and-2OOto +4OO "C Application.Rev. Sci. Instrum. 1970,41,1444-1446. Shen, K.-P.; Li, M.-H. Solubility of Carbon Dioxide in Aqueous Mixtures of Monoethanolamine with Methyldiethanolamine. J. Chem. Eng. Data 1992,37,96-100. Vickery, D. J.; Campbell, S. W.; Weiland, R. H. Gas Treating With Promoted Amines. Proceedings of the Laurance Reid Gas Conditioning Conference;University of Oklahoma: Norman, OK, 1988; pp 1-16. Weast, R. C., Ed. Handbook of Chemistry and Physics, 61st ed.; CRC Press: Boca Raton, FL, 1980. Zhang, D.; Razzaghi, M.; Ng, H.-J. Modelling of Acid Gas Treating With Blended Amines. Presented at the 72nd Annual Convention of the Gas Processors Association, San Antonio, TX, March 1517, 1993.

+

Received for review April 20, 1994 Accepted May 10, 1994a @

Abstractpublishedin Advance ACSAbstracts, June 15,1994.