Solvent Development for Aqueous Absorption/Stripping of CO2

The University of Texas at Austin J. Tim Cullinane and Gary T. Rochelle April 27, 2004

Outline
Overview Process Considerations Solvent Development
Experimental Methods Development of Aqueous K+/PZ

Other UT Research Activities

Degradation Process Modeling Pilot Plant/Packing Selection

Conclusions

U.S. CO2 Emissions from Fossil Fuel Combustion by Sector

Commercial 4.8% Industrial 31.8%

Residential 9.7%

Power Plant - Petroleum 2.0% Power Plant - Natural Gas 4.6%

Power Plant - Coal 47.1%

Total U.S. Emissions = 3635.7 Tg CO2 Eq.

Excludes Transportation, EPA (1999)

Advantages of Aqueous Absorption/Stripping

Near Commercial Technology
Process used for treating H2 & natural gas MEA demonstrated on small coal plants Promoted K2CO3 used for H2 treating

Post-process Technology Development

Lower cost and less risk to process Resolve problems in small pilot plants Demo Full-scale absorbers with 100 MW gas

Problems
20 - 40% energy use High capital cost

Enhancing CO2 Capture by Amines

1. Contactor Development
Packing Multi-pressure stripper Inter-cooling Power plant specific Large-scale equipment

2. Process Flowsheet Innovations

3. Energy Integration 4. Engineering Development 5. Solvent Development

CO2 Capture by Amines

Sweet Gas 1% CO2
PCO2* ~ 300 Pa

Cooler

2-4 mol H2O/mol CO2

Absorber T = 4060oC

Stripper T = 100120oC

PCO2* ~ 3000 Pa

Sour Gas 10% CO2

Rich Amine

Lean Amine

Reboiler H = 20-25 kcal/mol CO2

Solvent Development K+/PZ

1. Thermodynamics 2. Rates of Absorption 3. Degradation 4. System Modeling
Process Flowsheet

Bench-Scale Work

Fundamental

5. Pilot Plant

Large-Scale Work

CO2 Absorption by K+/Piperazine

Carbonate Species
O O O

O C O

H O H O

O OH O

Piperazine Species

H N

N H

C O

H N

N O

O H N N O H N N H O H N N O

O C O H H H H N
+

O N O N H O N O N

O O

PZ Speciation by

1H

NMR

H2 C O C N C O H2 H2 H2 O C C N C HN O C C H2 H2

H2 O C N C O C H2 H2 C HN C H2 H2 O C N C O C H2

H2 H2 C C NH HN C C H2 H2

Wetted-Wall Column
Pressure Control (35 60 psig)

IR CO2 Analyzer Condenser Pump (2 4 cm3/s)

WWC (38 cm2) Sample Port

N2
Heater Saturator oC) (25 110 Solution Reservoir o (25 110 C) (1000 cm3)

CO2

Flow Controllers (4 6 L/min)

Fundamental Equilibrium Modeling

Uses Electrolyte NRTL Model
Rigorous activity coefficient model

Benefits
Versatile can be used for broad range of conditions, systems Develops/supported by theory more accurate extrapolations Predicts complicated behavior

Challenges
Accurate representation of entire system Meaningful results can require a lot of data Thermodynamic consistency

Model Parameter Summary

System
K2CO3, H2O KHCO3, K2CO3, H2O PZ, H2O PZ, CO2, H2O PZ, K+, CO2, H2O

Data Types
Boiling pt. elev., PH2O* PCO2* UNIFAC NMR, PCO2* NMR, PCO2*

Parameters
4 4 4 3a 7

Data Points
681 120 21 406 203

a. 6 parameters for equilibrium constants also regressed

Speciation in 1.8 m PZ at 60oC

1.0

~300 Pa
0.9 0.8 PZ Total Reactive Species

~10000 Pa

Fraction of Total PZ

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2

PZH+ H PZCOO PZCOO

+ -

PZ(COO-)2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Loading (mol CO2/mol PZ)

Speciation in 5.0 m K+/2.5 m PZ at 60oC

1.0

~300 Pa
0.9 0.8 PZ

~10000 Pa
Total Reactive Species

Fraction of Total PZ

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0 0.1 0.2 0.3 0.4 0.5

PZCOO

PZH PZ(COO )2
-

H+PZCOO-

0.6
+

0.7

0.8

0.9

1.0

Loading (mol CO2/(mol K + mol PZ)

Equilibrium in K+/PZ at 60oC

PZ

/1 .8
+

m
6m PZ m 3.6 1.8

10000
K

PZ 2 6.

PCO * (Pa)

5.

100
7m

10

1 0 1 2 3 4

[CO2(aq)] Absorbed (m)

K
EA M ) % t 0w (3

/2 .5

1000

PZ

.2 /1

3. 6

/0.

PZ

Heat of Absorption
20

3000 Pa at 60 C
18
5.0 m K /2.5 m PZ
+

1.8 m PZ

-Habs (kcal/mol CO2)

16

14
6.2 m K+/1.2 m PZ

12
3.6 m K Model Predictions
+

10

Other Model Predictions + 3.6 m K Experimental Points 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

6 0.0

2*[PZ] 2*[PZ] + [K+]

Normalized Flux at 60oC

3.6 m K /3.6 m PZ Normalized Flux (mol/Pa-cm2-s) 5.0 M MEA 3.6 m K /1.8 m PZ
+ +

5.0 m K /2.5 m PZ

6.2 m K /1.8 m PZ
1e-10

3.6 m K /0.6 m PZ

2.5 m K /2.5 m PZ
100 1000 10000
2

PCO * (Pa)

Absorption Rate in 5.0 m K+/2.5 m PZ

1e-9

Normalized Flux (mol/cm2-Pa-s)

100oC 60oC 110oC

40oC
1e-10

80oC

100

1000

10000

PCO * (Pa)
2

Research Activities at UT
Bench-scale
Wetted-wall Column VLE, rates NMR speciation Degradation Other solid solubility, transport properties

Modeling
Thermodynamics Rate Process

Pilot Plant
Contactor Testing Solvent Testing

Oxidative Degradation of MEA

NH3
H HO CH2 CH2 N H

O2

Formaldehyde

Formate, Acetate Rate is measured by NH3 evolution from a sparged reactor vessel Gas analysis is quick/liquid analysis requires long experiments Uncertainty in the stoichiometry of O2 in the reaction

Degradation Results
Study
Rooney et al. Blachly and Ravner Girdler Hofmeyer et al. Chi and Rochelle Goff and Rochelle

Sparge Gas
Air Air 50% O2 Pure O2 Air Air Air w/ Agitation

Gas Flow/Liq. Max. Rate Vol (min-1) (mM/hr)

0.006 1.0 1.0 2.9 20.0 25.0 16.7 0.4 0.8 2.6 5.0 12.9 27.8 45.8

Conclusion: Mass Transfer Limited?

Process Modeling
Explore Optimum Operating Conditions
Heat requirement (kcal/gmol CO )
2

60 50 Optimal MEA

optimum lean

P*
CO2

1.25 kPa

2.5 40 40C Absorber 1.6 atm stripper 30 3 3.2 3.4 3.6 3.8 lean loading (m) 4 5 10 4.2

Process Configuration
Explore unique flowsheets
Multipressure Stripper
Rich ldg=0.46 Lean ldg=0.34 113 C 115 C 118 C 4 atm CO2 130 atm

2.8 atm Multistage Compressor W=7.4 kc/mol CO2 Q=20 kc/mol CO2 2 atm

Pilot Plant at UT-SRP

18 PVC columns, 20 ft of packing
Accommodates commercial, structured packing Operation as absorber or absorber/stripper

Vacuum stripping Air/CO2 Fed

Wide Range of Concentrations Possible

Pilot Plant Operation

NaOH/Air Screen packing areas
Packing areas based on 0.75 H2O/ft, 5 gpm/ft2

Packing
CMR 2, plastic IMTP #40 CMR 2, metal Montz B1-250 Montz B1-350

Wetted Area (ft2/ft3)

27 44 48 64 91

Solvent Simulation of absorber/stripper

Quantify real solvent performance Includes impurities (Fe2+, degradation, etc.)

Conclusions
E-NRTL model describes speciation and VLE K+ increases the amount of reactive species in solution
CO32-/HCO3- is an effective buffer Apparent carbamate stability is increased w/ K+

Solvent capacity increases with concentration and is comparable to MEA Habs can be lower than other amine-based systems and depends on the ratio of K+:PZ Absorption rate is 1.5 to 4 times faster than MEA or other amine-promoted K2CO3 solutions

Acknowledgements
Texas Advanced Technology Program: contract 003658-0534-2001 George Goff Degradation Tunde Oyenekan Process Modeling Dr. Ben Shoulders The University of Texas at Austin, Department of Chemistry

Questions?