Aucun titre de diapositive -...

Membranes and Post Combustion CarbonDioxide Capture: Challenges and Prospects

Eric FAVRE

ENSIC Nancy UniversitéLaboratoire Réactions & Génie des Procédés CNRS

NANCY FRANCE

i) Introduction: Membranes & Post combustion capture

ii) Membrane contactors

iii) Gas separation membranes

iv) Chemically reactive membranes and alternative approaches

vi) Conclusion

Outline

0 2 4 60

2

4

6

c2 (g

/L)

c1 (g/L)

SMB Classique M3C Continu M3C Moyen

Introduction

Separation processes in industry at a glance

I. Thermal energy

II. Auxiliary phase

III. Membrane

Separation agent:

(Humphrey & Keller, Separation Process Technology, 2005)

Van der Sluis, J.P. et al. C. Feasability of polymer membranes for carbon dioxide recovery from flue gases, Energy Conversion & Management, 33, 5-8 (1992) 429.

O. Davidson, B., Metz, Special Report on Carbon Dioxide Capture and Storage, International Panel on Climate Change , Geneva, Switzerland, 2005

Gas separation membranes and post combustion carbon capture

Source: Figueirao J. et al. DOE (2007) Int. J. Greenhouse Gas Control

Carbon capture processes: A tentative roadmap 1/2

Carbon capture processes: A tentative roadmap 2/2

Carbon capture strategy

Target mixture Conditions First generation separation

process

Possible breakthrough

membrane process

Oxycombustion O2/N2 P atmosphericT ambient

Cryogeny Ion Transfer Membranes (ITM)

Precombustion CO2/H2 P up to 80 BarT 300 – 500 C

Gas-liquid absorption in

physical solvent

Membrane reactor

Postcombustion CO2/N2 P atmosphericT 100 – 250 C

Gas-liquid absorption in

chemical solvent (MEA)

Membrane gas separation

Membranes & carbon capture processes

Membranes and post-combustion carbon capture: Synopsis

• Intensified gas liquid absorption: Membrane contactors

• Gas permeation membranes: physical

• Gas permeation membranes: chemical

• Alternative approaches & miscellaneous

Procédés de capture du CO2

i) Absorption gaz liquide conventionnelle

ii) Absorption par contacteurs à membranes

iii) Adsorption

iv) Membranes0 2 4 6

0

2

4

6

c2 (g

/L)

c1 (g/L)


Membranes contactors for intensified

CO2 absorption processes





• Alternative approaches, miscellaneous

CAPEX

Membrane contactors for intensified gas-liquid absorption: Principle

dP �� cos4max��

• Key issue: design a porous materialand use it between the liquid solvent and the gas phase under non wettingconditions

• Advantages:- increased interfacial area (a) - no flooding or weeping limitations- improved liquid distribution, no sensitivity to orientation- limited solvent losses- scale-up modular and easy

• Drawbacks:-additional mass transfer resistance- no economy of scale (numbering up)

Membrane contactors: a novel intensification process for fluid mass tranfer processes

Hollow fiber moduleInternal fiber diameter ~ 0.2-5 mmSpecific interfacial area (a): 1000-5000 m-1

� � � �Lmlm

e

eG

eINTG

kEmrkr

rkrK

..1

..

1,

�

�

�

Membrane contactors in industry at a glance

Membrane phase contactor unit for stripping of oxygen from water. (Provided by Hoechst Celanese Corporation).

• Numerous industrialapplications (carbonatedbeverages, aroma recovery, oxygen stripping…

• Dozens of publications each month on CO2 capture at lab scale…

• Very limited feedback and limited number of actors atpilot and industrial scale(Statoil, TNO)

• 2 main membrane materials: PP and PTFE

r

r e

MEA

solution

CO2mixture

Gas -liquid

interfacer

Liquid

boundary

layer

Gas

boundary

layer

Microporous

membrane

Liquidbulk Gasbulk

z

��

..m

mD

k �

Membrane contactors for CO2 capture: Materials challenge

• D: diffusion coefficient (bulk – Knudsen)

• �: Porosity

• : Membrane thickness

• �: Tortuosity

Trade-off between permeability and breakthrough pressure

Polypropylene (PP)

Polyvinylfluoridilène (PVDF)

Polytetrafluoroethylene (PTFE)

Nylon

Porous hydrophobic membrane materials: Examples of different structures

Membrane contactors and CO2 absorption in 30% MEA:A parametric study

Reference case: Packed column

Membrane contactor

Inlet solvent (MEA)

loading ( IN)

0.24

(Abu-Zahra et al., 2007)

Variable (usually 0)

Outlet solvent (MEA)

loading ( IN)

0.48

(Abu-Zahra et al., 2007)

Variable

Gas phase pressure drop

(�PG)

5000 – 7000 Pa

(NETL, 2007)

Usually not mentioned

Gas-liquid interfacial area

(a)

200-500 m-1

(Tobiesen et al., 2007)

1000 – 5000 m-1

(Gabelman & Hwang,

1999)

Overall mass transfer

coefficient (K)

4x10-3 – 10-2 m.s-1

(Tobiesen et al., 2007)

3x10-4 – 10-3 m.s-1

(Feron et al., 2002)

Overall volumetric CO2

absorption capacity (C)

1 mol CO2 .m-3.s-1

(Tobiesen et al., 2007;

NETL, 2007; Mangalapally

et al., 2010)

0.7 - 10 mol CO2 .m-3.s-1

(Nishikawa et al., 1995;

Feron and Jansen, 2002;

Yeon et al., 2003; Hoff et

al., 2004; deMontigny et

al., 2005)

Parametric study: Targets and framework

1. Compute, through a number of simulations, the volumetric absorption capacity of differentmembrane contactors (isothermal conditions, plug flow on gas side, liquid in, countercurrent mode, 30 % MEA, 15%CO2 in feedmixture, capture ratio 0.9)

2. Select the set of data with a maximal gaspressure drop of 50 mBar and liquid pressure drop of 1 Bar

3. Identify the requirements in terms of membrane permeance, fiber geometry and module packing fraction which enableintensification to be achieved

0

20

40

60

80

100

0 0,01 0,02 0,03u

G

Lab scale tests and simulations with membrane contactors

A typical lab scale set-up for membrane contactors testing

Prediction performances of simulations with 30 % MEA

Variable

Variation range

Unit

re

hollow fiber external

radius

10-4 - 3.10-3

[m]

�

membrane thickness

5.10-5 - 10-4

[m]

�

module packing

fraction

0.1 - 0.6

[-]

uG

interstitial gas velocity

0.1 - 5

[m.s-1]

km

membrane effective

mass transfer coefficient

10-2 - 10-5

[m.s-1]

Parametric sensitivity: variables and range of variation

0

1

2

3

4

5

6

7

8

9

0 2000 4000 6000 8000 10000 12000

Inte

nsifi

catio

n fa

ctor

(I)

Parametric study: summary of results

Approximately 11 000 simulations performed

An intensification factor larger than 1 isobtained within the �P limits for ca 300 cases, with module length between 0.1 and 2m

A maximal intensification factor of 8.5 seems to be attainable for a very limitednumber of cases

An intensification larger than 1 requires: km >7x10-4 m/s < 80 µm and re < 400 µm

A packing fraction of 0.5-0.6 is requiredfor an intensification factor larger than 8

Ranking of the simulation results in termsof intensification factor

Membrane contactors and CO2 capture: one step forward….

Long time scale results are required for material selection

Experimental results at a large scale, on real flue gases are of primiraly importance

Non isothermal conditions, water evaporation, gas and liquid dispersion effectsshould be added in simulation packages

Desorption thanks to temperature resistantMC should be more systematically explored

Dense skin membrane contactors in order to prevent liquid wetting problems




iii) Adsorption


0

2

4

6

c2 (g

/L)

c1 (g/L)


Post combustion capture by gas

separation membranes






CAPEX + OPEX

Dense membrane permeation

JA

Membrane thickness

Upstream pressurepfeed

Downstream pressurepperm

�

Component AComponent B

pfeed > pperm

(1) Sorption on upstream side(2) Diffusion down partial pressure gradient(3) Desorption on downstream side

Permeability of A � PA = DA SA

where DA = diffusion coefficient

SA = solubility coefficient

Selectivity

/A A A

A BB B B

P S DP S D

� � � � � �

� ��

Membrane Gas Separation: Applications & Market

0.1

1

10

100

1000

0.1 1 10 100 1000 104

*

Permeability (Barrers)

O2/N

2

CO2/N

2

CO2/CH

4

H2/N

2

Trade-off curves predictions for physical process

Step 1: Selectivity challenge

Feed composition xCapture ratio RPurity y

Selectivity challenge

Capture Cost

in

p

QQ

��

P’ : Upstream pressure

P’’ : Downstream pressureQp: Flowratey: CO2 mole fraction

Stage cut

PIn: Feed pressureQin: Flowratexin: CO2 mole fraction

RetentateFeed

Permeate To compression

Qout: Flowratexout: CO2 mole fraction

Key variables

Pressure ratio

2

2*N

CO

��

�

'''PP

��

Selectivity

Gas permeation module design: single stage case

Separation performances:

R: Recovery= Capture ratio

y: Permeate content CO2

Q in Q out x in x out

y Q p

dQ

Q Q - dQ x x - dx

P ’

P ’’

FEED RETENTATE

PERMEATE

z dA

Permeation module simulation: Cross plug flow case

ODE system resolution : DASSL (M3Pro© software)

��

��

dsdQxyx

dsdxQ p

*...* �

� ��

�� )1.(11.*

pp yxyxds

dQ �

�

��

��

�

�

)1.(1.

.1 p

p

p

p

yxyx

yy

��

Recovery (target > 80%)Pe

rmea

te c

onte

nt y

Bounaceur, R., Lape, N., Roizard, D., Vallières, C., Favre, E. (2005) Membrane processes for post-combustion carbon dioxide capture : a parametric study. Energy

Gas permeation module master / curves

10% CO2 in flue gas:

Impossible to achieve

R> 0.8 and y >0.8

unless a > 200

20% CO2 in flue gas:

Possible to achieve

R> 0.8 and y >0.8

if > 50

Gas permeation module master curves

Step 2: Energy challenge



Operating Conditions��

Energy Challenge

E < 2 GJth/tonOPEX

Energy cost

Capture Cost

Energy requirement computation: Feed compression

��

�

�

��

�

��

�

��

�

�

1PP

.1

RT..QE

1

min

maxIn1

��

��

!



PIn: 1 atmQin: Flowratexin: CO2 mole fraction

RetentateFeed

Permeate


0

2

4

6

8

10

12

0 50 100 150 200

Ener

gy E

(GJ /

ton

CO2 re

cove

red)

xin = 0.3

xin = 0.2

xin = 0.1 R = 0.8 y = 0.8

1. High parametric sensitivity: E strongly influenced by xin2. For a 10% CO2, content the energy requirement and the minimal

membrane selectivity are too high

Parametric study results: Feed compression

Energy requirement computation: Vacuum pumping

��

�

�

��

�

��

�

��

�

�

1PP

.1

RT..'

QE

1

min

maxP1

��

��

!



PIn: 1 atmQin: Flowratexin: CO2 mole fraction

RetentateFeed

Permeate Pipe & injection


0

0.5

1

1.5

2

2.5

3

0 50 100 150 200

Ene

rgy

E (

GJ

/ ton

CO

2 re

cove

red)

xin = 0.1

xin = 0.2

xin = 0.3

R = 0.8 y = 0.8

Parametric study results: Vacuum pumping

1. E below 0.5 GJ per ton can be achieved 2. For a 20% CO2, content the required membrane selectivity

for E = 0.5 GJ per ton is around 60, i.e. a realistic figure

The ultimate (and largely unexplored) challenge: energy, productivity, cost



Operating Conditions��

Energy Challenge

E < 2 GJth/ton

Feed flow ratePermeabilityThickness

ProductivityChallenge

Maximal flux

OPEX

CAPEX

Energy cost

Membrane cost(€/m2)

Capture Cost

Tackling the energy (pressure ratio) / capacity trade off

Pressure ratio

Capture ratio

Surface area

15% CO2 in feed mixture 90% purity at permeate

Ind. Eng. Chem. Res. (2008) 47, 5, 1562

Main conclusion:

Best solution : single stage withvacuum

Selectivity: > 40 required (but y

will be limited to 0.6)

Ind. Eng. Chem. Res. (2008) 47, 5, 1562

Main conclusion (ctd):

High permeability required (500 Barrer with z 0.125 µm)

Ind. Eng. Chem. Res. (2008) 47, 5, 1562

Main conclusion (ctd):

Low membrane cost required (< 50 $/sqm)

* CO2/N2maximum ~ 60

� CO2maximum ~ 500 Barrer

Selectivity challenge: "Physical" separation membranes

Favre, E. (2007) Carbon dioxide recovery from post combustion processes: Can gas permeation membranes compete with absorption? Journal of Membrane Science, 294, 50-59

MTRPolaris membrane

~ 501000 GPU

The only way to overcome the energy challenge seems to apply vacuum, nevertheless, this option may show major limitations:

- energy efficiency can be low compared to compressors (! ~ 0.85)- vacuum pump foot print is large and leaks can be problematic

Option 1: Single stage membrane system

Herzog et al., Environ. Prog. (1991) 10, 64-74.

CO2 recovery 80%, CO2 purity 90%Energy requirement 50-75 % of combustion energy of coal

(MEA 47-79 %)

One of the earliest membrane flowsheet for post-combustion CCS application

Option 2: Multi-stage membrane systems

Challenges & unsolved issues: Water

1

10

100

1000

10000

0,1 1 10 100 1000 10000

P CO2

(Barrer)

CO2 / N

2

CO2 / Ar

CO2 / O

2

CO2 / SO

2

CO2 / NO

CO2 / CO

0,1

1

10

100

0,1 1 10 102 103 104 105

CO

2/O

2

CO2 permeability (Barrer)

Challenges & unsolved issues: Oxygen and “minor species”

Rejection increases




iii) Adsorption


0

2

4

6

c2 (g

/L)

c1 (g/L)


MMembranes and post

combustion carbon capture:

Chemically reactive membranes






CAPEX + OPEX

Looking for improved performances : Chemically reactive membranes with fixed sites

• Polyelectrolyte membranes (PVBTAF) with water saturatedmixture permeation: = 900 (Quinn,1977)

• Plasma grafted polyacrylic acidmembrane with amine carrier (Matsumya, 1994) = 4700, P = 106 Barrer!!

• Polyvinylamine (Hagg, 2005) = 150-250

• PAMAM (Duan, 2006) , = 260

Looking for improved performances : Liquid membranes

• Numerous studies for a long time!

• WWard & Robb (1966) > 4600

• SSirkar, Ho, Noble…

Membrane type Material and/or carrier CO2/N2selectivity

CO2 permeability (Barrer) or permeance

(GPU)

Gas separation membrane (dense

polymers)

PEO-PBTPEG/Pebax©

PEG-DME/ Pebax©

PEGDA/PEGMEAPolaris™

7047434150

120 Barrer151 Barrer600 Barrer570 Barrer1000 GPU

Fixed Site Carrier

Membrane(FSCM)

PAAM-PVA / PSPVAm/PVAPEI / PVAPDMA/PSPDMAMA

801452305380

24 GPU212 GPU

1 GPU30 GPU5 GPU

Liquid Membrane

(LM)

PVAm-PVA/PSPVAm/PVAAmines/PVA

Carbonic anhydraseAmines / PVA

9090500250493

22 GPU15 GPU250 GPU80 GPU

693 Barrer

Chemically reactive membranes: Performances

Assymetric PVTMS membrane selective layer thickness 0.2 µm

gas-penetrant

gas-carrier

liquid carrier260 µm

membranes

Carrier K2CO3 aqueous solution

Liquid membrane processes

Shalygin, M.G., Roizard, D. Favre, E., Teplyakov V.V. (2008). Journal of Membrane Science, 318, 317-326.

PVTMS

PTMSP

PE PDMSEthyl

cellulose

PP

Water

Ethanol Methanol

Propylenecarbonate

1,0E+00

1,0E+01

1,0E+02

1,0E+03

1,0E+04

1,0E+01 1,0E+02 1,0E+03 1,0E+04 1,0E+05CO2 permeability coefficient, Barrer

CO

2/N

2 s

elec

tivi

ty

Selectivity challenge: "Chemical" separation membranes

Fixed site carrier membranes:Selectivity 80-250Water required

Liquid membranes:Selectivity up to 4500Water required

Key issues:Productivity (l ~5-7 µm)Driving force (sweep)Stability

Carbonic anhydrase LM: Carbozyme concept




iii) Adsorption


0

2

4

6

c2 (g

/L)

c1 (g/L)


Membranes processes for CO2 capture:

Alternative approaches

Looking for alternative emission sources…

��

��

��"�

yyyyy

xxxxxTW )-1ln().-1()ln(. -)-1ln().-1()ln(..-min

0

5000

10000

15000

0 0,2 0,4 0,6 0,8 1

W

(J.mol -1)

x (% vol)

y = 1

y = 0.9

y = 0.7

y = 0.5

AirGas Coal Steel,

cementFermentation

NH3, EO

Wmin

x

y

Towards breakthrough membrane materials?

Zeolite membranes performances. Predictions based on MD simulations

(J. Memb. Sci., in press)

Unconventional approach : Reverse selective membranes

CO2 selective membrane N2 selective membrane

y

RFavre, E., Roizard, D., Koros, W.J. (2009) Ind. Eng. Chem. Res., 48 (7) 3700-3701.

Hybrid processes

+=

x)x-1ln().x-1()xln(.x

T.R-W 'min

��

��

�"��

��

"�y

)y-1ln().y-1()yln(.y .Tx

)x-1ln().x-1()xln(.xT.-W ''min

2

1

Air Power

3

CompressioTransport Storage

4

y x’ x

E1 E2

N2 �� O2

O2 N2

CO2 H2O N2

Fuel N2 �� CO2

Flue gas

CO2

H2O

CO2 N2

Hybrid process: Rationale

Cryogeny Gas permeation membrane module

Combustion Flue gas drying

44'.x.R10.1

"P'P.

'T.

1n21n3.

RE

2E1EE3

1

2O

��

�

�

��

�

�

��

��"

�

��

��

!��

Hybrid process:Principle

Natural gas (n =1)Most challenging!

Single module

Feed compression

R = 0.9

y = 0.9

Hybrid process: Simulation results

�50-60% oxygen in feed air seems to be the best compromise

Favre, E., Bounaceur, R., Roizard, D. (2009). Separation & Purification Technology, 68, 30-36.

0,0

0,5

1,0

1,5

2,0

2,5

3,0

3,5

4,0

4,5

5,0

0 0,2 0,4 0,6 0,8 1

y

E [

GJ/

ton

CO

2]

XCH4 = 0.95

XCH4 = 0.90

XCH4 = 0.80

XCH4 = 0.70

XCH4 = 0.50R = 0.9y = 0.9 = 50

Carbon dioxyde capture by a membrane unit from biogas combustion unit : simulation results

Oxygen mole fractionin enriched air

Methane mole fractionin biogas

Favre, E., Bounaceur, R., Roizard, D. (2009) Biogas, membranes and carbon dioxide capture. Journal of Membrane Science, 328, 11-14.

Miscellaneous….

• Flue gas drying

•Compressor purge

•Alernate driving forces: temperature, wator vapor sweep

• Pressurised combustion (e.g.Combicap concept)

• Integrated approach (S, N, C removal)

• High temperature permeation

….




iii) Adsorption


0

2

4

6

c2 (g

/L)

c1 (g/L)


Conclusion

• Membranes processes offer a large variety of potential applications in a CCS framework (intensification, separation, concentration)

• For the capture step, the purity and recovery specifications play a key role(strong parametric sensitivity, strong flexibility)

•Material science & process engineering collaboration is crucial!

• Pilot studies & technico-economical analyses are needed!

•Post combustion CCS is a complex, multivariable landscape: no silver bullet!

•Gas separation membranes missed (as usual) the first generation

Gas permeation membranes and post combustion CCS: utopy or opportunity?

Acknowledgements:

N. Boucif, E. Chabanon, P.T. Nguyen

R. Bounaceur, C. Castel, S. Rode, D. Roizard

Support & funding:

CNRS

ANR

EU / FP7 (CESAR)

Further information?

Questions?

[email protected]

Target Membrane process Main characteristics

Intensified gas liquid absorption

Membrane contactor(hydrophobic porous membrane)

Membrane stability (non wetting conditions) has to be ensured for long time operationRegeneration step non achievableEffective intensification factor remains to be clearly evaluated

Carbon dioxide capture

1. Gas separation membranes: physical (GS) (usually a dense polymer)

Numerous materials investigated, few technico-economical studies, almost no pilot scale process on real flue gasWater in flue gas could be a problemRequire multistage processes in order to achieve purity target

2. Chemically reactive membranes (FSCM)

Require water, on both sides of the membrane,to be effectiveMay be applicable through a single stage process if a high selectivity is attained Industrial feedback on real flue gas still not achieved

3. Liquid membranes (LM) Very high selectivity can be achievedRequire water, on both sides of the membraneto be effectiveSolvent losses due to volatility can be a problem, possibly solved by ionic liquidsEffective drivig force, stability and effective permeability in a real psot-combustion situation remain to be

Liquid nitrogen

Adsorption

Cryogeny

Membrane

N2 purity

Flowrate (Nm3.h-1)10 250 5000

90

95

99.9

Source: Greenhouse Issues (2006), 84, 12

Source: Greenhouse Issues (2006), 84, 12

Date post:	03-Feb-2020
Category:	Documents
Upload:	others
View:	4 times
Download:	0 times

Aucun titre de diapositive -...

Documents