Solubility of K2SO4 in CO2 Loaded MEA/PZ Solution

Jan 10th, 2008Qing Xu

Rochelle groupDepartment of Chemical Engineering

University of Texas at Austin

Outline Reclaiming process

Why reclaimingReclaiming processes

Experiment Conductivity for K2SO4 solutionRanges of the experimentsApparent Ksp dependence on T, amine and I

Regression Analysis – Electrolyte-NRTL (in Aspen Plus 2006.5) Interaction parameters regressionFlash simulation test for regression results

Conclusion Future work

Why reclaiming? What to reclaim:

SO2, HCl, NOx SO4-2, Cl-, NO3

-…

Organic acids, amine polymers Why reclaiming?

To remove heat stable salts and degradation products, which may cause corrosion, foaming, and will affect the capacity of amine solution.

Chemical reactions:

MEACOOMEAHCO2MEA

)(SO)(MEAHOH

O2

1SO2MEA

2

242

222

Reclaiming Processes:

Thermal reclaiming with addition of NaOH Na2SO4(s)

MEA degradation products Sodium formate

Electrodialysis Ion exchange Sulfate precipitation

Fertilizer

MEASOK)SO(MEAHKOH 4224

ABSORBER

STRIPPER

SCRUBBER

FLUEGAS

CACO3

CASO3

PUREGAS

RICH-L

RICH-H

CO2

LEAN-H

LEAN-L

CO2 Capture ProcessHigh T

Low T

CaSO4

CaCO3

FLASH2

HEATER

CRYSTALLIZER FILTER

COMPR

Precipitation Reclaiming for Hot Lean Solvent

3×

Hot Lean

Cool Lean

CO2

H2O

K2SO4(s)KOH

Solubility of K2SO4 in CO2 Loaded Amine can Determine:

Solvent flow rate for reclaiming.

Operating condition to avoid scaling.

Energy and equipment cost.

Apparatus and Method - 1

Stirrer

K2SO4

Conductivity meter

Water Bath

Apparatus and Method - 2

BottleTop

Water Bath

ConductivityMeter

Water Bath

y = 99.4

y = 36.584x + 81.254

95

95.5

96

96.5

97

97.5

98

98.5

99

99.5

100

0.35 0.40 0.45 0.50 0.55 0.60

[K2SO4](m)

co

nd

(mS

/cm

)

11m MEA, [CO2]t=0.4 mol/mol MEA, 80oC

Conductivity~[K2SO4]

Start with K2SO4 slurry

End, clear solution

Ranges of the Experiments

T: 25 , 40 , 80 ℃ ℃ ℃ Amine: 7m MEA, 11m MEA, 4m PZ, 7m MEA/2m

PZ, 8m PZ, 10m PZ, etc. CO2 loading:

Loading

Extra K+: 0, 0.35m Extra SO4

=: 0, 0.15m

0.5 0.4, 0.2, 0.1, 0.05, 0,α PZ of moles*2MEA) of (moles

loaded CO of molesα 2

0.001

0.01

0.1

1

0.8 0.9 1 1.1 1.2 1.3 1.4I0.17

Ksp

3.75-T

1966.4ine]0.36[eq.amI9.15lnKsp 0.17

)m],SO([)m],K([Ksp 24

2

n

1i

2iizc

2

1I

Ksp dependence on T, amine concentration, and ionic strength

7m

11m

80C40C

Regression Analysis - Ksp

Ksp of K2SO4: Property method: elecNRTL in Aspen Plus 2006.5 Based on solubility data of K2SO4 in water by Söh

nel, 1985, regress A, B, and C in Ksp(T):

Parameter Value (SI units) σ

A 235.0 2.5

B -13227 118

C -36.2 0.4

ln(T)*CB/TA(T) Kspln

)γ(SO)x(SO)(Kγ)(KxKsp 24

24

22

Originally by Chen et al., for aqueous electrolyte systems. Later extended to mixed solvent electrolyte systems (Mock et al., 1984, 1986).

A versatile model for the calculation of activity coefficients.

Adjustable ENRTL interaction parameters: Molecule-molecule Molecule-electrolyte Electrolyte-electrolyte Each interaction consists of both the nonrandomness

factor and energy parameters .

Regression Analysis – Electrolyte NRTL Model

Regression Analysis – Interaction parameters

Framework Data Regression System in Aspen Plus Electrolyte-NRTL model

Chemistry: Ksp(T) from previous regression.

]T

Tln

T

TT[E

T

DC

ref

ref

B,caB,ca

B,caB,ca

298.15KT

a andc pair eelectrolyt—ca

molecule solvent—B

ref

]T

Tln

T

TT[E

T

DC

ref

ref

ca B,ca B,

ca B,ca,B

｝ T

DC

Existing parameters Electrolyte-NRTL default values in Aspen Plus Hilliard, 2005 (10 C, D & E for MEA-H2O-CO2 system and parameters f

or pure/binary components)

Objective Parameters (C) and (D) related to K+ and SO4

=

Solubility data used Water data by Söhnel, 1985; MEA data by Xu, 2006~2007.

Method Select parameters with small correlation coefficients. Small residual root mean square error. Small AARD (average absolute relative deviations) in flash simulation t

est

Regression Analysis – Interaction parameters

Regression Analysis - Results

Component i Component j C DH2O (K+,MEACOO-) x x

(K+,MEACOO-) MEA x x(MEA+,SO4

-2) H2O x x(MEA+, SO4

-2) MEA xMEA (MEA+,SO4

-2) x(K+, SO4

-2) MEA x x(K+, HCO3

-) MEA x x

A series of flash simulation Using the regressed interaction parameter set Simulate under each experimental condition Get activity coefficient and mole fraction for each

case, calculate Ksp and the error from Ksp(T).

AARD: 45.6%

Regression Analysis – Test

)T(Ksp

)SO()K(

)T(Ksp

)SO()SO(x)K()K(x

SALT)-K Ksp(from

t)coefficienactivity andfraction Ksp(from

)T(Ksp

)meas(Ksp

-24

224

24

22

)SO()x(SO)K()K(xKsp 24

24

22

Dependence of Ksp(meas)/Ksp(T) on [CO2]

0.1

1

10

0 2 4 6

[CO2](m)

Ksp

(me

as)

/Ksp

(T)

Dependence of Ksp(meas)/Ksp(T) on [MEA]

0.1

1

10

0 5 10[MEA](m)

Ks

p(m

ea

s)/

Ks

p(T

)

Dependence of Ksp(meas)/Ksp(T) on T

0.1

1

10

20 40 60 80

Temperature(C)

Ksp

(me

as)

/Ksp

(T)

Conclusions Ksp dependence on T, equivalent amine concentr

ation, and ionic strength:

A parameter set in Electrolyte-NRTL model for the CO2-MEA-H2O-K+-SO4

= system was developed. Still need modification.

Ksp(meas)/Ksp(T) is dependent on MEA concentration and CO2 loading, independent on temperature.

3.75-1966.4

]eq.amine[63.09.15ln 17.0

TIKsp

Future work Modify the regression; get more accurate

C/D/E parameter set for the . Conduct experiments of K2SO4 crystallization

Slurry characteristics: settling rates, filterability, drying rates and final moisture contents.

Crystal characteristics: composition, form, habit, shape factors, and solid density.

Crystallization kinetics: crystal nucleation, growth.

Modify Aspen reclaiming process model with regressed parameters, optimize conditions, add crystallization data into the model.