1

A Membrane Process to Capture CO2from Coal-Fired Power Plant Flue Gas

Projects NT43085 and NT05312

T. Merkel, H. Lin, X. Wei, J. He, B. Firat, K. Amo, R. Daniels, R. BakerMembrane Technology and Research, Inc.

NETL review meetingMarch 26, 2009

www.mtrinc.com

2

• Project timeline and objectives overview

• Introduction to MTR

• Membrane technology background

• Results to date (project NT43085)

• New project (NT05312) objectives and future plans

Outline

3

Project number NT43085Project period: 4/1/07 to 3/31/09 Funding: $788,266 DOE; $197,066 MTR

DOE program manager: Heino Beckert, Participants: MTR, DOEBruce Lani

Project scope: investigate the feasibility of new polymer membranes and process for cost-effective capture of CO2 from power plant flue gas.

All project objectives were met within time and budget; details to follow.

Project number NT05312Project period: 10/1/08 to 9/30/10 Funding: $3,439,200 DOE; $957,630 cost share

DOE program manager: Jose Figueroa Participants: MTR, APS, EPRI, DOE

Project scope: field demonstrate the MTR membrane process with commercial-sized components at APS’s Cholla coal-fired power plant; at the conclusion of the project, be in a position to gauge the technical and economic viability of membrane-based CO2 capture from flue gas.

Project overview

4

Natural Gas:

Petrochemicals:

Hydrogen (Refinery):

Propylene/Nitrogen

CO2/CH4, CH4/N2NGL/CH4

H2/CH4, CO, CO2

Introduction to MTR

MTR designs, manufactures, and sells membrane systems for industrial gas separations

5

10-3

10-2

10-1

100

101

102

103

104

0 100 200 300

Permeability (Barrer)

PolydimethylsiloxaneRubbery Polymer

PolysulfoneGlassy Polymer

H2

O2

N2

CO2

CH4

C2H6

C3H8

C2Cl2F4

CCl2F2

SF6

CH4

N2

O2

CO2

NH3

H2He

Gas critical volume (cm3/mol)

Membrane technology basics

⎟⎠

⎞⎜⎝

⎛×⎟

⎠

⎞⎜⎝

⎛==α

B

A

B

A

B

AB/A D

DSS

PP

Solubility selectivity: favors large molecules

Diffusivity selectivity: favors small molecules

1 Barrer = 10-10 cm3 (STP) cm / (cm2 s cmHg)

6

• Spiral-wound and hollow fiber modules are used.

• Membranes have to be thin to provide useful fluxes.

Membrane technology basics

7

Flue gas cleanup in a coal-fired power plant

BoilerCoal

CO2

AirESP FGD

Ash

Steam to turbines

Membrane challenges for treating this large volume of low-pressure gas:• Large membrane area needed → high CO2 permeance is a must!• How to generate driving force w/o using large compression or vacuum power• Potential harmful contaminants include fly ash, SO2, water, and trace metals.

Sulfur

600 MW =500 m3/s flue gas = 1540 MMscfd = 460 tons CO2/h

8

Polaris™ membranes are extremely permeable to CO2

10

20

30

40

50

60

100 1,000 10,000

CO2/N2selectivity

CO2 permeance (gpu)

PolarisTM

Target area identified from

design calculations

Commercial CA membranes

Pure-gas data at 25°C and 50 psig feed pressure

9

• Countercurrent sweep with combustion air provides “free” driving force.

• 90% capture as liquid CO2 can be achieved using about 15% of plant energy.

• Even with Polaris™, 0.5 to 1 million m2 of membrane area are required for 90% CO2 capture from a 600 MWe plant.

MTR’s process design for flue gas CO2 capture

10

High CO2 permeance reduces capture cost more than high selectivity

10

15

20

25

30

35

40

0 20 40 60 80 100 120

Cost of capture

($/ton CO2)

Membrane CO2/N2 selectivity

1,000 gpu

1,500 gpu

3,000 gpu

90% CO2 recovery

PolarisTM 1

PolarisTM 3

CO2Permeance

Cost assumes membrane at $150/m2 and includes CO2 compression

11

Polaris™ membrane and module scale-up• More than 500 m2 of membrane were produced on commercial-sized equipment.

• 8-inch diameter modules were fabricated and tested.

Polaris™ membrane has been scaled-up to commercial size

12

Mixed-gas module performance

10

100

1,000

20 30 40 50 60

Mix

ed-g

as p

erm

eanc

e (g

pu)

System temperature (oC)

CO2

N2

Mixed-gas selectivity is lower than pure-gas and decreases with increasing temperature. At 50°C, module selectivity is still >20.

10

20

30

40

50

20 30 40 50 60

Mix

ed-g

as C

O2/N

2 sel

ectiv

ity

System temperature (oC)

module 5414

Permeance Selectivity

13

Spiral-wound countercurrent sweep modules work

The optimal capture process uses about 100% sweep/feed ratio. At this sweep rate, the module shows a 10-fold increase in CO2 flux compared to no sweep.

0.1

1

10

0 30 60 90 120 150Sweep/Feed Flow Rate (%)

CO2 flux

(slpm/m2)

Feed = 10% CO2 in N2

Measured

Theoretical

No Sweep

14

APS Red Hawk power plant• Uses 8-inch diameter Polaris™ membrane modules.

• Captured CO2 will be used for biofuels production.

Polaris™ membrane systems have proven field experience

BP Pascagoula • Conducted 3-month field test of 8-inch diameter

Polaris™ membrane modules.

• Removed CO2 and higher hydrocarbons from raw natural gas.

15

Cholla field test

• A 6-month field test of 8-inch diameter Polaris membrane modules at Cholla coal-fired plant is scheduled to begin in 4Q09.

• Key objectives are to demonstrate sweep operation in commercial-sized modules and to investigate membrane lifetime.

Skid footprint is 24’ x 7’250,000 scfd flue gas slipstreamCaptures 1 ton CO2/day

16

• Total energy use in plant is 56 MW; uses 5.5 MW water pumps.

• Plant produces 100 million m3/yr of fresh water.

Membrane plants of the required size exist today

Ashkelon desalination plant• 40,000 spiral-wound RO

membrane modules

• 1.5 million m2 membrane area

17

Efficient module packing and low-cost components reduce installed cost

About 130 mega-module skids would be required for a 600 MWe power plant.

Typical cost ($/m2) Type of unit High-pressure gas

separation (steel vessels) Reverse osmosis (plastic vessels)

Membrane 20 5

Membrane modules 100 10-15

Installed modules in a skid 570 20-50

18

Advantages and challenges

• Particulate matter and its potential impact on membrane lifetime• Energy losses due to feed and permeate side pressure drops• Cost reduction and device scale-up issues• Integration with power plants and effect of recycle on boiler• Lack of operating experience in the power industry

• Simple operation; no chemical reactions, no moving parts

• Tolerance to high levels of wet acid gases; inert to oxygen

• Compact and modular with a small footprint

• Low energy use; no additional water used (recovers water from flue gas)

• Builds on existing, low-cost technology already used at a similar scale

Advantages:

Challenges:

19

Future plans beyond the current project

Develop plans for a 50-100 ton CO2/day system

Include CO2 liquefaction section

Incorporate low-cost components (target $150/m2)

Explore options for testing air recycle to boiler

20

Summary

• Earlier project demonstrated the potential of high permeance membranes and a novel process design to cost-effectively capture CO2 from coal-fired flue gas.

• Operating experience with real flue gas is critical. The upcoming field test at Cholla will be a key test of the membrane approach.

• The scale-up and cost reduction roadmap can be modeled on existing membrane technology.

• Next step is a 10-fold larger system that demonstrates low-cost components and CO2 liquefaction.

21

Acknowledgements

• DOE NETL – Heino Beckert, Jose Figueroa• APS – Ray Hobbs, Xiaolei Sun• EPRI – Abhoyjit Bhown, Brice Freeman, Dick Rudy,

George Offen

22

Extras

23

24

Why doesn’t high membrane selectivity help?

• high selectivity demands high membrane area.

• increasing the pressure ratio is energy intensive.

0

20

40

60

80

100

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

Permeate CO2

concentration(%)

Permeate pressure (bar)

Feed = 13% CO2 at 1.1 bar

α = 50

membrane selectivity(α ) = 500

Minimum practical permeate pressure(pressure ratio =10)

Understanding membrane process design is important to direct materials development

The separation is pressure ratio limited, and

25

The importance of pressure ratio

For this example,

• the concentration of CO2 in the permeate can never exceed 50% no matter how selective the membrane

• At least half the permeate must be the slow component (N2)

• Permeation of the slow component determines the membrane area

• Infinite selectivity = no slow component permeation = infinite membrane area

Permeate0.2 bar; 50% CO2

PCO2 = 50% of 0.2 bar = 0.1 bar

Feed1 bar; 10% CO2

PCO2= 10% of 1 bar = 0.1 bar

50% 110% 0.2

feed

permeate

Ppermeateconcfeed conc P

barbar

⎛ ⎞⎛ ⎞ ≤ ⎜ ⎟⎜ ⎟⎝ ⎠ ⎝ ⎠

⎛ ⎞ ⎛ ⎞≤⎜ ⎟ ⎜ ⎟⎝ ⎠ ⎝ ⎠

Membraneenrichment

Pressureratio≤

26

Would a membrane with higher selectivity help?

0

20

40

60

80

100

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

Permeate CO2

concentration(%)

Permeate pressure (bar)

Feed = 13% CO2 at 1.1 bar

α = 50

α = 500

Minimum practical permeate pressure(pressure ratio =10)

No! The separation is pressure ratio limited, and high selectivity demands high membrane area

0

5

10

15

20

10 100 1000

Relative membrane

area

Membrane selectivity (CO 2/N2)

Feed = 13% CO2 at 1.1 barpermeate = 0.11 barpressure ratio = 10

27

Polaris™ membranes are extremely permeable to CO2

10

20

30

40

50

60

100 1,000 10,000

CO2/N2selectivity

CO2 permeance (gpu)

Polaris 1

Polaris 2

Polaris 3

Target area identified from

design calculations

Commercial CA membranes

Pure-gas data at 25°C and 50 psig feed pressure

28

• Initial 2-year project (4/1/07 to 3/31/09) investigated the feasibility of new polymer membranes to cost-effectively capture CO2 from flue gas.

• All objectives were met within time and budget; key findings include:

– Membranes can be fabricated into commercial modules; membrane permeance is 10x higher than existing materials and stability in acid gases looks good.

– For a real-world membrane system, membrane permeance is much more important than selectivity.

– Novel design shows promise to capture 90% CO2 using <15% of plant energy.

• New 2-year, $4.4 million project (10/1/08-9/30/10) with EPRI and APS will field demonstrate the membrane process with commercial-sized components; key objectives are:

– Run a 6-month field test at APS’s Cholla coal-fired power plant.

– With EPRI, conduct a comparative economic analysis of the proposed process.

– Develop low-cost component prototypes and a cost reduction roadmap.

– Investigate scale-up issues and begin plans for next-stage, 10x larger demonstration.

Project overview

29

Key remaining challenges to be addressed in the current project

• Particulate matter and its potential impact on membrane lifetime

• Energy losses due to feed and permeate side pressure drops

• Cost reduction and device scale-up issues

• Lack of operating experience