Experimental Vapor–Liquid Equilibrium Data for Binary Mixtures of

6 days ago - Methyldiethanolamine (MDEA) is a widely used chemical in the natural gas processing industry as a solvent for CO2 and H2S capture and as ...
1 downloads 4 Views 1MB Size
Download PDF
Article pubs.acs.org/jced

Cite This: J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Experimental Vapor−Liquid Equilibrium Data for Binary Mixtures of Methyldiethanolamine in Water and Ethylene Glycol under Vacuum Adam Soames,*,† Ammar Al Helal,† Stefan Iglauer,‡ Ahmed Barifcani,† and Rolf Gubner† †

WA School of Mines: Minerals, Energy and Chemical Engineering, Curtin University, Bentley Western Australia 6102, Australia Petroleum Engineering Department, Edith Cowan University, Joondalup Western Australia 6027, Australia

‡

S Supporting Information *

ABSTRACT: Methyldiethanolamine (MDEA) is a widely used chemical in the natural gas processing industry as a solvent for CO2 and H2S capture and as a basic compound for pH stabilization corrosion control. During pH stabilization corrosion control, the removal of MDEA during the (mono)ethylene glycol (MEG) regeneration process may occur under vacuum conditions during reclamation in which the removal of salt cations is performed. Isobaric vapor−liquid equilibrium data for the binary MEG−MDEA system is presented at (20, 10 and 5) kPa and water−MDEA system at (40, 20, 10) kPa to simulate its behavior during MEG reclamation under vacuum. Vapor and liquid equilibrium concentrations of MDEA were measured using a combination of ion chromatography and refractive index. The generated experimental VLE data were correlated to the UNIQUAC, NRTL, and Wilson activity coefficient models, and the respective binary parameters were regressed.

■

INTRODUCTION Ethylene glycol (EG) and N-methyldiethanolamine (MDEA) are common chemicals used in the natural gas processing industry. The injection of EG is performed to prevent the formation of natural gas hydrates within transportation pipelines.1−3 Whereas MDEA and other alkanolamines are typically used as chemical absorbents for the removal of carbon dioxide and hydrogen sulfide during natural gas processing.4,5 Furthermore, the application of MDEA within natural gas transportation extends to its use as a basic compound suitable for pH stabilization corrosion control.2,6 pH stabilization corrosion control is performed to promote the formation of an iron carbonate protective film by artificially increasing the system pH.2,6−8 MDEA as a pH stabilizer may be preferable to salt based (hydroxide or carbonate) chemicals because of its ability to be recovered during vacuum reclamation minimizing operational losses and dosing requirements.2,9 Moreover, the thermal stability of MDEA is advantageous during industrial EG regeneration where exposure to high temperature (120−140 °C)3 is required allowing multiple regeneration cycles before thermal degradation occurs.10,11 Vacuum reclamation is often performed to prevent the accumulation of salts within the EG regeneration loop.12−14 The vacuum reclamation process entails the vaporization of EG to remove nonvolatile salt compounds. Vacuum reclamation of EG is typically performed at low pressure (≈100 mbar12,15,16) to minimize the required operational temperature (120−150 °C13,16). Low temperature vaporization of EG is desired to prevent its degradation.14,15 However, the vacuum reclamation process may © XXXX American Chemical Society

inadvertently lead to MDEA losses due its higher boiling point in comparison to EG. Therefore, ensuring the vaporization of MDEA alongside EG is an important aspect of EG regeneration during pH stabilization to minimize MDEA losses. Alternatively, the removal of MDEA within EG regeneration systems operating under pH stabilization control is essential following formation water breakthrough.14,15 The combined presence of MDEA (high pH) and divalent cations including calcium, magnesium, and barium presents a scaling risk within both transportation lines and equipment operating at high temperature (heat exchangers, EG regeneration system).14,17 MDEA will react in the presence of CO2 to form bicarbonate2,14,17 facilitating the formation of scaling products including CaCO3. pH stabilization chemicals such as MDEA must therefore be removed to facilitate switch over to more scaling friendly film forming corrosion inhibitors (FFCIs). The removal of MDEA can be achieved via vacuum reclamation systems alongside mineral salts.14 Therefore, knowledge of the vapor−liquid equilibrium (VLE) of MDEA with respect to EG at low pressure is essential for the design of separation equipment. Current literature for MDEA VLE data in EG and water solutions is limited at the low-pressure conditions necessary for EG vacuum reclamation. This work outlines the VLE of MDEA with respect to water and EG under low pressure conditions (40−10 kPa) and (20−5 kPa), respectively. However, the operating conditions of reclamation Received: January 16, 2018 Accepted: March 21, 2018

A

DOI: 10.1021/acs.jced.8b00054 J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data

Article

systems for EG regeneration may ultimately depend on whether MDEA removal of retention is desired.

■

EXPERIMENTAL METHODOLOGY Materials. MDEA and EG were purchased from Chem Supply with a mass purity greater than 99.5% wt and were used without further purification. Where water was used, deionized water with a resistivity of 18.2 MΩ.cm was utilized. The physical properties of water, EG, and MDEA including refractive index of the pure solutions, their boiling points, and critical properties are outlined in Table 1 and Table 2.

Table 1. Chemicals, Suppliers, and Purity chemical

supplier

CASRN

purity (mass %)

ethylene glycol methyldiethanolamine water

Chem Supply

107-21-1 105-59-9

>99.5 >99 18.2 MΩ·cm

Apparatus and Procedure. The VLE data of the MEG− MDEA system under vacuum was generated using a Heidolph

Figure 2. Refractive index vs MDEA mole fraction calibration curve at 20 °C, 101.325 kPa (water−MDEA).

Table 2. Refractive Indices (nD), Boiling Points, and Critical Properties refractive indices (nD) at 20 °C, 101.325 kPa

a

chemical

lit.

measureda

boiling point (°C) at 101.325 kPa

Tc/K

Pc/MPa

Zc

water ethylene glycol methyldiethanolamine

1.3325 1.431819,20 1.464224

1.3323 1.4315 1.4684

100 197.321 24725

647.09618 64522,23 67822,23

22.06418 8.57322 3.8822

0.2318 0.26222 0.25422

Standard uncertainties are u(nD) = 0.0003, u(T) = 0.1 °C and u(P) = 0.1 kP.

Figure 1. Experimental apparatus (Heidolph Hei-VAP Rotary Evaporator). B

DOI: 10.1021/acs.jced.8b00054 J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data

Article

Table 3. continued x2 0.926 0.965 0.991 1.000

y2 0.046 0.120 0.415 1.000

γ1 10 kPa 0.825 1.001 1.213

γ2

T/K

0.773 0.924 1.037 1.000

385.68 400.89 425.54 448.42

a

Standard uncertainties are u(T) = 0.1 K, u(P) = 0.1 kPa, and u(x,y) = 0.005.

Table 4. VLE Data and Calculated Activity Coefficients (γ) for MEG−MDEA Systema x2

Figure 3. Refractive index vs MDEA mole fraction calibration curve at 20 °C, 101.325 kPa (EG-MDEA).

Table 3. VLE Data and Calculated Activity Coefficients (γ) for Water−MDEA Systema x2

y2

0.000 0.356 0.532 0.756 0.825 0.880 0.912 0.928 0.939 0.950 0.966 0.980 1.000

0.000 0.0015 0.0035 0.013 0.023 0.042 0.065 0.087 0.109 0.142 0.221 0.359 1.000

0.000 0.075 0.224 0.295 0.446 0.613 0.675 0.785 0.856 0.912 0.955 0.980 1.000

0.000 0.0001 0.0005 0.0008 0.0017 0.0038 0.0054 0.012 0.022 0.045 0.112 0.260 1.000

0.000 0.188 0.276 0.436 0.565 0.659 0.766 0.826 0.876

0.0000 0.0003 0.0005 0.0011 0.0022 0.0037 0.0075 0.013 0.022

γ1 40 kPa 1.000 0.993 0.992 0.932 0.950 0.977 1.009 1.006 1.001 1.037 1.034 1.107 20 kPa 1.000 1.016 1.010 1.006 0.931 0.914 0.893 0.862 0.902 0.955 1.031 1.184 10 kPa 1.000 0.972 0.948 0.954 0.864 0.798 0.772 0.755 0.776

γ2

T/K

1.101 1.042 0.925 0.924 0.939 0.939 0.934 0.930 0.968 0.941 1.000 1.000

349.02 360.12 368.57 388.95 398.56 409.65 418.65 425.32 430.65 435.92 447.59 458.69 488.42

1.197 1.151 1.133 0.991 0.913 0.881 0.824 0.832 0.856 0.922 0.982 1.000

333.20 334.56 338.57 340.80 348.20 357.50 362.50 374.50 384.56 397.56 414.80 432.60 467.58

1.111 1.057 1.012 0.885 0.782 0.752 0.725 0.742

318.95 323.65 326.50 331.60 339.48 346.89 356.80 364.91 373.21

y2

0.000 0.031 0.088 0.135 0.283 0.385 0.434 0.491 0.572 0.621 0.676 1.000

0.000 0.004 0.013 0.022 0.059 0.102 0.127 0.154 0.210 0.256 0.315 1.000

0.000 0.041 0.078 0.110 0.204 0.366 0.460 0.610 0.668 0.755 0.832 1.000

0.000 0.005 0.011 0.016 0.036 0.092 0.140 0.239 0.300 0.413 0.541 1.000

0.000 0.039 0.118 0.201 0.276 0.363 0.453 0.598 0.682 0.732 0.835 0.925 1.000

0.000 0.005 0.017 0.034 0.052 0.082 0.117 0.210 0.283 0.345 0.515 0.725 1.000

γ1 20 kPa 1.000 0.993 0.998 0.986 0.963 0.956 0.939 0.928 0.893 0.872 0.851 10 kPa 1.000 0.992 0.990 0.992 0.978 0.940 0.926 0.888 0.862 0.808 0.789 5 kPa 1.000 0.984 0.990 0.995 0.999 0.968 0.951 0.931 0.887 0.878 0.860 0.841

γ2

T/K

0.763 0.755 0.790 0.814 0.854 0.855 0.882 0.913 0.930 0.992 1.000

423.01 423.56 424.81 426.25 430.89 434.25 436.37 438.65 442.60 445.11 447.50 467.46

0.620 0.687 0.727 0.842 0.980 1.002 0.986 1.000 0.992 0.988 1.000

406.01 406.65 407.50 408.15 410.67 415.69 418.83 425.15 428.03 433.13 437.49 448.32

0.804 0.860 0.873 0.911 0.918 0.915 0.928 0.953 0.981 0.975 0.997 1.000

390.40 391.10 392.40 394.00 395.90 398.79 401.78 407.14 411.27 413.51 419.26 424.90 430.52

vacuum rotary evaporation system. The system is designed to perform vacuum distillation and was modified to permit the generation of VLE data using the flow-scheme shown in Figure 1. The system is capable of generating a vacuum down to 20 mbar with an accuracy of ±1 mbar. Furthermore, the system is capable of producing a maximum solution temperature within the heating flask of 180 °C ± 0.1 °C using an oil heating bath. The system was modified from its original design by placing a manually adjustable valve between the condenser and condensed C

DOI: 10.1021/acs.jced.8b00054 J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data

Article

equation fitted to the respective calibration curve. Furthermore, nitrogen was continuously introduced into the rotary flask to prevent the thermal degradation of MDEA and MEG in the presence of oxygen. Thermal degradation would otherwise result in discolouration that would impact the refractive index measurement.26−28

liquid storage to continuously return the condensed vapor to the flask. Liquid reflux of the condensed vapor was maintained until a constant temperature of the liquid and vapor phases was achieved indicating equilibrium. Following temperature stabilization, the valve was opened to allow accumulation of the condensed vapor with the storage vessel. Liquid and vapor phase samples were extracted from the system and analyzed by a combination of ion chromatography and refractive index measurement. Ion chromatography was performed using a Dionex ICS2100 IC System. For comparison purposes, the mole fraction of MDEA was also measured via refractive index at 20 °C using an ATAGO PAL-BX/RI refractometer with an accuracy of ±0.0003. The calibration curve of nD vs mole fraction (MDEA) is given by Figure 2 and Figure 3. Tabulated refractive index data are available in the Supporting Information. The MDEA concentration was calculated using the nD of the sample and a polynomial

■

RESULTS AND DISCUSSION Experimental Results. The experimental VLE data of the water−MDEA at P = (40, 20, 10) kPa and EG-MDEA P = (20, 10, and 5) kPa binary systems is presented in Table 3 and Table 4, respectively. To assess the applicability of using the modified rotary evaporator for generation of VLE data, the water−MDEA VLE data at 40 kPa were compared to literature data reported by Voutsas et al.29 with good agreement found in (Figure 4).

Figure 7. VLE data for MEG−MDEA at 20 kPa. Figure 4. VLE data for water−MDEA at 40 kPa.

Figure 5. VLE data for water−MDEA at 20 kPa.

Figure 8. VLE data for MEG−MDEA at 10 kPa.

Figure 6. VLE data for water−MDEA at 10 kPa.

Figure 9. VLE data for MEG−MDEA at 5 kPa. D

DOI: 10.1021/acs.jced.8b00054 J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data

Article

Table 5. Water Vapor Pressure and EG Antoine Parameters water eq 2 parameters

a1

a2

a3

a4

a5

a6

−7.85951783

1.84408259

−11.7866497

22.6807411 B

−15.9618719

1.80122502

A EG Antoine parameters34

8.0908

C −67.70

2088.94 log(psi /mmHg) = A −B/[T/K − C ]

Furthermore, comparison was also made to the limited water− MDEA VLE data reported by Barreau et al.30 for water−MDEA at 10 kPa with good agreement again found (Figure 6). Likewise, the VLE data generated for the MEG−MDEA system was compared to the literature data reported by Yang et al.22 (Figure 7). Because of the temperature limitations of the rotary evaporator system, VLE data up to approximately 175 °C was generated for the MEG−MDEA system at 20 kPa. The liquid-phase activity coefficients γ for each chemical was calculated from the experimental data by eq 1,31 where psi represents the vapor pressure of the pure component i at equilibrium temperature.22 As the VLE data were generated at low pressure, the behavior of the vapor phase can be considered ideal and the φ factor considered negligible.22,32 γi =

Figure 10. MEG−MDEA δT residual.

yP i

s ϕi

xipi

(i = 1, 2) (1)

Estimation of water vapor pressure was performed using the empirical correlation proposed by Wagner and Pruß33 as per eq 2. The correlation provides accurate estimation of water vapor pressure over a wide range of temperatures (273.15−633.15 K). Conversely, estimation of ethylene glycol vapor pressure was achieved using the Antoine coefficients listed in Table 5.34 For MDEA, the vapor pressure was calculated via the Clausius− Clapeyron type equations proposed by Voutsas et al.29 and Xu et al.35 given by eqs 3 and 4, respectively. The equation used to calculate MDEA vapor pressure was dependent on equilibrium temperature with the applicable temperature ranges of eqs 3 and 4 being 413−513 K29 and 323−383 K,35 respectively. Figure 11. MEG−MDEA δy1 residual.

⎛P ⎞ T ln⎜ water ⎟ = C (a1θ + a 2θ1.5 + a3θ 3 T ⎝ Pc ⎠ + a4θ 3.5 + a5θ 4 + a6θ 7.5)

(2)

ln ps = 26.1369 −

⎛ 7588.516 ⎞ ⎜ ⎟, ⎝ ⎠ T

(3)

ln ps = 26.2942 −

⎛ 7657.86 ⎞ ⎜ ⎟, ⎝ T ⎠

where θ=1−

T TC

ps = Pa ps = Pa

(4) Figure 12. MEG−MDEA δ(γ1/γ2) residual.

Table 6. Herrington Thermodynamic Consistency Test system

pressure (kPa)

D−J

water−MDEA

40 20 10 20 10 5

−31.62 −26.59 −27.91 −4.01 −4.38 −5.12

EG-MDEA

Thermodynamic Consistency. The experimental VLE data were analyzed using the semiempirical thermodynamic consistency test for isobaric binary VLE data proposed by Herington.36 For isobaric VLE data, experimental data can be considered thermodynamically consistent when (D − J) is less than 10.22,36,37 The variables D and J were evaluated using eqs 5 and 6, respectively, where the values of “area+” and “area−” are calculated from the x1−ln(γ1/γ2) graph. The calculated D and E

DOI: 10.1021/acs.jced.8b00054 J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data

Article

composition, x1 to determine if the deviations scatter uniformly about zero.32 If the VLE data are thermodynamically consistent, the residual plots should show no clear trend else systemic errors may be present.32 The residual plots of δT, δy1, and δ ln(γ1/γ2) for the MEG−MDEA system is illustrated by Figures 10−12, showing the experimental data are thermodynamically consistent. The water−MDEA system showed similar consistency. Correlation of VLE Data. Correlation of the experimental VLE data reported was performed using the NRTL, Wilson, and UNIQUAC models. Regression of the respective binary parameters was performed using the objective function outlined by eq 722,32 where γexp was calculated through eq 1. The models and their respective equations are outlined by Table 10 with the utilized UNIQUAC parameters reported in Table 7. For the NRTL model, the nonrandomness parameter αij was set to 0.3.22,38 The regressed binary parameters for each model are presented in Table 8 and Table 9 for water−MDEA and MEG− MDEA, respectively. The comparison of experimental data to calculated values is illustrated by Figures 4 to 6 for the water− MDEA system and Figures 7 to 9 for the MEG−MDEA system.

J values for each system are summarized in Table 6 confirming thermodynamic consistency. D=

(area + ) − (area − ) 100 (area + ) + (area − )

J = 150

(5)

Tmax − Tmin Tmin

(6)

Furthermore, the thermodynamic consistency test proposed by Van Ness31 was also applied by comparing the predictions of the nonrandom two liquid (NRTL),38 Wilson,39 and universal quasichemical (UNIQUAC)40 models to the experimental data. The model residuals were plotted against the liquid-phase Table 7. Molecule Volume Parameters r, Area Parameters q, and Z Parameter for the UNIQUAC Model chemical

r

water EG MDEA

29,40

0.92 2.408022,34 4.9441022,29

q

Z

29,40

22,40

1.40 2.24822,34 4.26822,29

10

2

⎛γ − γ ⎞ exp cal ⎟ OF = ∑ ⎜⎜ ⎟ γ ⎝ ⎠ exp

Table 8. Water−MDEA Binary Interaction Parameters Water−MDEA system pressure

40 kPa

20 kPa

NRTL 84.37 −135.34 −197.94 −119.53 Wilson 4856.33 4880.69 −4289.02 −4872.10 UNIQUAC 332.95 169.77 594.60 527.19

gij − gjj/J·mol−1 gji − gii/J·mol−1 gij − gjj/J·mol−1 gji − gii/J·mol−1 gij − gjj/J·mol−1 gji − gii/J·mol−1

10 kPa

To assess the accuracy of the regressed binary parameters and corresponding model fit, the root-mean-square (RMS) deviations δT, δy1 δ ln(γ1/γ2) were calculated via eq 8.32 The respective RMS values are presented in Tables 11 and 12 for

98.10 −919.05 4378.64 −4998.23

Table 11. Water−MDEA RMS Error for Model Fitting water−MDEA

−332.11 357.00

system pressure

MEG−MDEA 20 kPa NRTL −921.85 −338.24 Wilson −783.00 1894.99 UNIQUAC 40.15 72.33

gij − gjj/J·mol−1 gji − gii/J·mol−1 gij − gjj/J·mol−1 gji − gii/J·mol−1 gij − gjj/J·mol−1 gji − gii/J·mol−1

10 kPa

5 kPa

−566.61 −662.28

−545.48 −371.01

−743.31 1547.58

−654.29 1757.05

50.24 70.65

40.00 65.11

40 kPa

RMS δT/K RMS δy1 RMS δ ln(γ1/γ2)

Table 9. MEG−MDEA Binary Interaction Parameters system pressure

(7)

RMS δT/K RMS δy1 RMS δ ln(γ1/γ2) RMS δT/K RMS δy1 RMS δ ln(γ1/γ2)

NRTL 0.829 0.024 0.155 Wilson 0.719 0.037 0.190 UNIQUAC 1.298 0.053 0.251

20 kPa

10 kPa

0.810 0.021 0.160

0.990 0.036 0.259

0.812 0.029 0.170

0.899 0.036 0.259

1.301 0.050 0.260

1.354 0.064 0.373

the water and MEG MDEA systems. Comparative RMS values to the work of Kim et al.32 and Wang et al.41 were calculated in

Table 10. Activity Coefficient Models NRTL

⎡ ⎤ τjiGji2 τijGij2 ⎥ ln γi = xj⎢ + 2 2 ⎢⎣ (xi + xjGji) (xj + xiGij) ⎥⎦

Wilson

⎛ Aij Aji ⎞ ⎟ ln γi = − ln(xi + Aijxj) + xj⎜⎜ − xj + xiAji ⎟⎠ ⎝ xi + xjAij

UNIQUAC

⎛ ⎛θ ⎞ ⎛ϕ⎞ ⎛Z⎞ r ⎞ ln γi = ln⎜ i ⎟ + ⎜ ⎟qi ln⎜⎜ i ⎟⎟ + ϕj⎜⎜li − i l j⎟⎟ − qi ln(θi + θτ j ji) rj ⎠ ⎝ xi ⎠ ⎝ 2 ⎠ ⎝ ϕi ⎠ ⎝ ⎛ τji τij ⎞ ⎟⎟ + θjqi⎜⎜ − θ + θτ θ + θτ ⎝ i j ji j i ij ⎠ F

DOI: 10.1021/acs.jced.8b00054 J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data

Article

Figure 15 compare the relative volatility calculated via the reported experimental MEG−MDEA VLE data and those

Table 12. MEG−MDEA RMS Error for Model Fitting MEG−MDEA system pressure

40 kPa NRTL 0.178 0.003 0.034 Wilson 0.203 0.006 0.054 UNIQUAC 0.345 0.017 0.072

RMS δT/K RMS δy1 RMS δ ln(γ1/γ2) RMS δT/K RMS δy1 RMS δ ln(γ1/γ2) RMS δT/K RMS δy1 RMS δ ln(γ1/γ2)

20 kPa

10 kPa

0.272 0.006 0.069

0.265 0.006 0.045

0.338 0.004 0.046

0.336 0.007 0.056

0.478 0.016 0.111

0.463 0.014 0.109

terms of δy1 − δ ln(γ1/γ2) and δT − δy1, respectively. The UNIQUAC model was found on average to give a higher RMS value, particularly in terms of δT/K suggesting the NRTL and Wilson models provide a better fit. The lower accuracy of the UNIQUAC model may be a result of the two regressed parameters being insufficient to accurately model the experimental data.42 n

RMS δX =

∑ i=1

(Xi − Xcalc)2 n

Figure 15. Comparison of experimental to modeled relative volatility for MEG−MDEA at 5 kPa.

calculated through the respective activity coefficient models. The deviation of the modeled results in comparison to experimental data appears to be greatest at the lower MDEA concentration region as suggested by Mathias.43 This is primarily due to the low concentration of MDEA within the vapor phase (y2 < 0.01) where the reported measurement uncertainty (u(x,y) = 0.005) can lead to a significant change in calculated relative volatility. The percentage error in modeled MDEA K-values with respect to the experimental data is illustrated by Figure 16

(8)

To further assess the validity of the model fits, the VLE analysis method suggested by Mathias43 was applied. Figure 13 to

Figure 13. Comparison of experimental to modeled relative volatility for MEG−MDEA at 20 kPa.

Figure 16. Percentage error in model K-values in comparison to experimental data for MEG−MDEA.

where good agreement between experimental and NRTL and Wilson models was found. The absolute error for the NRTL and Wilson models was typically below 2%. Similar agreement was also found for the reported water−MDEA experimental data within the higher vapor−MDEA concentration regions (x2 > 0.4) (Figure 17). However, the effect of the measurement uncertainty was found to be more pronounced for the water− MDEA system at low vapor MDEA concentrations. Although a significant fraction of MDEA was present within the liquid phase, only minimal amounts were present within the vapor due to the extremely high relative volatility of water to MDEA.

Figure 14. Comparison of experimental to modeled relative volatility for MEG−MDEA at 10 kPa. G

DOI: 10.1021/acs.jced.8b00054 J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data

Article

(3) Zaboon, S.; Soames, A.; Ghodkay, V.; Gubner, R.; Barifcani, A. Recovery of mono-ethylene glycol by distillation and the impact of dissolved salts evaluated through simulation of field data. J. Nat. Gas Sci. Eng. 2017, 44, 214−232. (4) Closmann, F.; Nguyen, T.; Rochelle, G. T. MDEA/Piperazine as a solvent for CO2 capture. Energy Procedia 2009, 1, 1351−1357. (5) Arcis, H.; Rodier, L.; Ballerat-Busserolles, K.; Coxam, J.-Y. Modeling of (vapor+liquid) equilibrium and enthalpy of solution of carbon dioxide (CO2) in aqueous methyldiethanolamine (MDEA) solutions. J. Chem. Thermodyn. 2009, 41, 783−789. (6) Olsen, S.; Halvorsen, A. M. K. Corrosion Control by pH Stabilization. In Corrosion; NACE International: Dallas, TX, 2015. (7) Dugstad, A.; Seiersten, M.; Nyborg, R. Flow Assurance of pH Stabilized Wet Gas Pipelines. In Corrosion; NACE International: San Diego, CA, 2003. (8) Mendez, C.; Singer, M.; Comacho, A.; Nesic, S.; Hernandez, S.; Sun, Y.; Gunaltun, Y.; Joosten, M. W. Effect of Acetic Acid, pH and MEG on the CO2 Top of the Line Corrosion. In Corrosion; NACE International: Houston, TX, 2005. (9) Dugstad, A.; Seiersten, M. pH-stabilisation, a Reliable Method for Corrosion Control of Wet Gas Pipelines. In 1st International Symposium on Oilfield Corrosion, SPE 87560, , 2004. (10) Closmann, F.; Rochelle, G. T. Degradation of aqueous methyldiethanolamine by temperature and oxygen cycling. Energy Procedia 2011, 4, 23−28. (11) Voice, A. K.; Closmann, F.; Rochelle, G. T. Oxidative Degradation of Amines With High-Temperature Cycling. Energy Procedia 2013, 37, 2118−2132. (12) Brustad, S.; Løken, K. P.; Waalmann, J. G. Hydrate Prevention using MEG instead of MeOH: Impact of experience from major Norwegian developments on technology selection for injection and recovery of MEG. In Offshore Technology Conference, Houston, TX, 2005. (13) Bikkina, C.; Radhakrishnan, N.; Jaiswal, S.; Harrington, R.; Charlesworth, M. Development of MEG Regeneration Unit Compatible Corrosion Inhibitor for Wet Gas Systems. In SPE Asia Pacific Oil and Gas Conference and Exhibition, Society of Petroleum Engineers: Perth, Australia, 2012. (14) Latta, T. M.; Seiersten, M. E.; Bufton, S. A. Flow Assurance Impacts on Lean/Rich MEG Circuit Chemistry and MEG Regenerator/Reclaimer Design. In Offshore Technology Conference, Houston, TX, 2013. (15) Lehmann, M. N.; Lamm, A.; Nguyen, H. M.; Bowman, C. W.; Mok, W. Y.; Salasi, M.; Gubner, R. Corrosion Inhibitor and Oxygen Scavenger for use as MEG Additives in the Inhibition of Wet Gas Pipelines. In Offshore Technology Conference-Asia, Kuala Lumpur, Malaysia, 2014. (16) Psarrou, M. N.; Jøsang, L. O.; Sandengen, K.; Østvold, T. Carbon Dioxide Solubility and Monoethylene Glycol (MEG) Degradation at MEG Reclaiming/Regeneration Conditions. J. Chem. Eng. Data 2011, 56, 4720−4724. (17) Babu, D. R.; Hosseinzadeh, M.; Ehsaninejad, A.; Babaei, R.; Kashkooli, M. R.; Akbary, H. Carbonates precipitation in MEG loops − A comparative study of South Pars and Bass Strait gas fields. J. Nat. Gas Sci. Eng. 2015, 27 (2), 955−966. (18) Guideline on the Use of Fundamental Physical Constants and Basic Constants of Water; G5-01; IAPWS, 2016. (19) Fogg, E. T.; Hixson, A. N.; Thompson, A. R. Densities and Refractive Indexes for Ethylene Glycol-Water Solutions. Anal. Chem. 1955, 27, 1609−1611. (20) Kozma, I. Z.; Krok, P.; Riedle, E. Direct measurement of the group-velocity mismatch and derivation of the refractive-index dispersion for a variety of solvents in the ultraviolet. J. Opt. Soc. Am. B 2005, 22, 1479−1485. (21) Trimble, H. M.; Potts, W. Glycol-Water Mixtures Vapor Pressure-Boiling Point-Composition Relations. Ind. Eng. Chem. 1935, 27, 66−68. (22) Yang, C.; Feng, Y.; Cheng, B.; Zhang, P.; Qin, Z.; Zeng, H.; Sun, F. Vapor−Liquid Equilibria for Three Binary Systems of N-

Figure 17. Percentage error in model K-values in comparison to experimental data for water−MDEA.

The resulting low concentration of MDEA, coupled with the relatively high measurement uncertainty at such low concentrations, leads to the large error in calculated K-values.

■

CONCLUSION VLE for the water−MDEA and MEG−MDEA binary systems has been generated experimentally under low pressure conditions (40−5) kPa. Activity coefficients and binary interaction parameters for the binary systems were subsequently fit to the NRTL, Wilson, and UNIQUAC models. The experimental data were analyzed using the Herington36 and Van Ness31 thermodynamic consistency tests and found to be consistent. Comparison was also made to the limited MDEA VLE data reported by Voutsas et al.,29 Yang et al.,22 and Barreau et al.30 with good agreement found. The VLE data generated within this study is applicable for low pressure separation of MDEA from water and MEG solutions such as vacuum reclamation during industrial MEG regeneration.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jced.8b00054. Refractive index of the water−MDEA and EG−MDEA solutions (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Adam Soames: 0000-0001-6883-2231 Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Haghighi, H.; Chapoy, A.; Burgess, R.; Tohidi, B. Experimental and thermodynamic modelling of systems containing water and ethylene glycol: Application to flow assurance and gas processing. Fluid Phase Equilib. 2009, 276, 24−30. (2) Pojtanabuntoeng, T.; Kinsella, B.; Ehsani, H.; McKechnie, J. Assessment of corrosion control by pH neutralisation in the presence of glycol at low temperature. Corros. Sci. 2017, 126, 94−103. H

DOI: 10.1021/acs.jced.8b00054 J. Chem. Eng. Data XXXX, XXX, XXX−XXX

Journal of Chemical & Engineering Data

Article

(43) Mathias, P. M. Guidelines for the Analysis of Vapor−Liquid Equilibrium Data. J. Chem. Eng. Data 2017, 62, 2231−2233.

Methylethanolamine, N-Methyldiethanolamine, and Ethylene Glycol at P = (40.0, 30.0, and 20.0) kPa. J. Chem. Eng. Data 2013, 58, 2272− 2279. (23) Yaws, C. L. Ed. Chemical Properties Handbook: Physical, Thermodynamic, Environmental, Transport, and Health Related Properties for Organic and Inorganic Chemicals; McGraw-Hill Education LLC.: New York, N.Y., 1999. (24) Zhang, Q.; Cai, S.; Zhang, W.; Lan, Y.; Zhang, X. Density, viscosity, conductivity, refractive index and interaction study of binary mixtures of the ionic liquid 1−ethyl−3−methylimidazolium acetate with methyldiethanolamine. J. Mol. Liq. 2017, 233, 471−478. (25) Lide, D. R. CRC Handbook of Chemistry and Physics, 88th ed.; Taylor & Francis, 2007. (26) AlHarooni, K.; Barifcani, A.; Pack, D.; Gubner, R.; Ghodkay, V. Inhibition effects of thermally degraded MEG on hydrate formation for gas systems. J. Pet. Sci. Eng. 2015, 135, 608−617. (27) Haque, M. E. Ethylene Glycol Regeneration Plan: A Systematic Approach to Troubleshoot the Common Problems. J. Chem. Eng. 2012, 27, 21−26. (28) Nazzer, C. A.; Keogh, J. Advances in Glycol Reclamation Technology. In Offshore Technology Conference, Houston, TX, 2006. (29) Voutsas, E.; Vrachnos, A.; Magoulas, K. Measurement and thermodynamic modeling of the phase equilibrium of aqueous Nmethyldiethanolamine solutions. Fluid Phase Equilib. 2004, 224, 193− 197. (30) Barreau, A.; Mougin, P.; Lefebvre, C.; Luu Thi, Q. M.; Rieu, J. Ternary Isobaric Vapor−Liquid Equilibria of Methanol + NMethyldiethanolamine + Water and Methanol + 2-Amino-2-methyl1-propanol + Water Systems. J. Chem. Eng. Data 2007, 52, 769−773. (31) Van Ness, H. C. Thermodynamics in the treatment of vapor/ liquid equilibrium (VLE) data. Pure Appl. Chem. 1995, 67, 859−872. (32) Kim, I.; Svendsen, H. F.; Børresen, E. Ebulliometric Determination of Vapor−Liquid Equilibria for Pure Water, Monoethanolamine, N -Methyldiethanolamine, 3-(Methylamino)-propylamine, and Their Binary and Ternary Solutions. J. Chem. Eng. Data 2008, 53, 2521−2531. (33) Wagner, W.; Pruß, A. The IAPWS Formulation 1995 for the Thermodynamic Properties of Ordinary Water Substance for General and Scientific Use. J. Phys. Chem. Ref. Data 2002, 31, 387−535. (34) Garcia-Chavez, L. Y.; Schuur, B.; de Haan, A. B. Liquid−liquid equilibrium data for mono ethylene glycol extraction from water with the new ionic liquid tetraoctyl ammonium 2-methyl-1-naphtoate as solvent. J. Chem. Thermodyn. 2012, 51, 165−171. (35) Xu, S.; Qing, S.; Zhen, Z.; Zhang, C.; Carroll, J. J. Vapor pressure measurements of aqueous N-methyldiethanolamine solutions. Fluid Phase Equilib. 1991, 67, 197−201. (36) Herington, E. F. G. Tests for the Consistency of Experimental Isobaric Vapor-Liquid Equilibrium Data. J. Inst. Pet 1951, 37, 457− 470. (37) Wisniak, J. A new test for the thermodynamic consistency of vapor-liquid equilibrium. Ind. Eng. Chem. Res. 1993, 32, 1531−1533. (38) Renon, H.; Prausnitz, J. M. Local compositions in thermodynamic excess functions for liquid mixtures. AIChE J. 1968, 14, 135−144. (39) Wilson, G. M. Vapor-Liquid Equilibrium. XI. A New Expression for the Excess Free Energy of Mixing. J. Am. Chem. Soc. 1964, 86, 127−130. (40) Abrams, D. S.; Prausnitz, J. M. Statistical thermodynamics of liquid mixtures: A new expression for the excess Gibbs energy of partly or completely miscible systems. AIChE J. 1975, 21, 116−128. (41) Wang, H.-x.; Xiao, J.-j.; Shen, Y.-y.; Ye, C.-s.; Li, L.; Qiu, T. Experimental Measurements of Vapor−Liquid Equilibrium Data for the Binary Systems of Methanol + 2-Butyl Acetate, 2-Butyl Alcohol + 2-Butyl Acetate, and Methyl Acetate + 2-Butyl Acetate at 101.33 kPa. J. Chem. Eng. Data 2013, 58, 1827−1832. (42) Kontogeorgis, G. M.; Folas, G. K. Activity Coefficient Models Part 2: Local Composition Models, from Wilson and NRTL to UNIQUAC and UNIFAC. In Thermodynamic Models for Industrial Applications; John Wiley & Sons, Ltd: 2009; pp 109−157. I

DOI: 10.1021/acs.jced.8b00054 J. Chem. Eng. Data XXXX, XXX, XXX−XXX