www.monash.edu.au

CO2 capture by Adsorption Processes: From Materials to Process Development to Practical Implementation

Paul A. WebleyDept. of Chemical EngineeringTalloires, July 2010

www.monash.edu.au

2

Outline

• Adsorption for Post-Combustion CO2

Capture – PVSA or TSA or TVSA?

• Materials development – synergy with

process and adsorbent assessment

• Lab and Pilot scale Post Combustion

Experimental Work

• Field Demonstrations

• Conclusions

www.monash.edu.au

3

Post Combustion Flue Gas Conditions and Challenges – T,P,Flow

PC(black)+FGD PC(brown) NGCC OxyfuelMass Flow (kg/s) 600 790 1000 180Volume flow (m3/hr)

2.2 x 106 3.0 x 106 3.8 x 106 0.5 x 106

Pressure (barg) ~ 0.05 ~ 0.05 ~ 0.05 ~ 0.05Temperature(°C) ~ 90 ~ 90 ~ 90 ~ 170

• Volume flows VERY large (pressure drop)• Pressures very low (no driving force) • Temperatures relatively high (for adsorption)

www.monash.edu.au

4

Post Combustion Flue Gas Conditions and Challenges - Composition

• CO2 concentration ranges from 12-63% (wet basis)

• High water content • SOx, NOx, ash, heavy metals, etc present

PC(black)+FGD PC(brown) NGCC OxyfuelNitrogen 71 60 75Carbon Dioxide 12.6 12 3.4 62.6Water 11.1 24 6.9 31.5Oxygen 4.4 3 13.8 4.5SO2 ~200ppm ~200ppmNOx 670 ppm 25ppm

www.monash.edu.au

5

Commercial Adsorption based CO2 Systems

• CO2 Recovery from COREX Gas–Saldanha Steel, South Africa

• 360,000 Nm3/h feed rate (15psig, 30%CO2)

• 2xVPSA trains• Each has 12 vessels

and 4 vacuum pump sets

• Worlds largest VPSA

Feed ~ 10,000 TPD

www.monash.edu.au

6

CO2 Capture: Program at Monash University

• Materials development:

– Good working capacity & good selectivity– Insensitive to moisture– Operable at above ambient temperature– There is a strong relationship to the process cycle

• Process development

– Cycle development– Structured adsorbents– Energy integration

• Solution of engineering issues

www.monash.edu.au

7

PSA or TSA?

• Throughput (kg CO2/day/kg adsorbent)

• TSA provides large WC at the expense of cycle time unless innovative rapid cycling systems are developed

• PSA provides smaller WC but can provide much shorter cycle time

• Large throughput, bulk separation tends to favor PSA, eg. O2VSA, H2PSA etc.

• PTSA??? (yes)

www.monash.edu.au

8

Adsorbent Requirements from Process Operation

• Minimal compression of flue gas! – vacuum operation => need large WC between 0 and 1 atm

• Regeneration of the adsorbent is where the energy is needed – put energy into CO2, not N2, not solid material (minimal N2 adsorption, minimal T swing)

• Difference between adsorption and desorption for CO2 compared to other gases is key

• Large adsorption amount is not necessarily better• Interaction of species is important (impurities)

www.monash.edu.au

9

Adsorbent Isotherms

CO2 Adsorption Isotherms at 25°C

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

4.50

5.00

0 20 40 60 80 100 120

P(kPa)

n(m

ol/

kg

)

13X

NaY

CaCHA

SWNT

BPL carbon

ACFAlumina

IRMOF-6

Feed

Reg

en

13X

WC,CO2 ~ 1.6

Isothermal Op.

www.monash.edu.au

10

Screening Problem

• Which adsorbent is “best” over a particular

operating region and what is this operating

region (can we solve the inverse problem)?

• By how much will improvements in adsorbent

reduce capture cost (or particular interest to

sponsors!)?

• Should improvements be aimed at CO2

capacity, N2 capacity, heat of adsorption, etc?

• Cannot simulate processes for every possible

adsorbent – too time-consuming.

www.monash.edu.au

11

Quick Assessment – “First Order” effects

• Need to consider adsorption P,T and desorption P,T of CO2 required to meet purity constraints

• Need to consider adsorption P,T and desorption P,T of N2 (selectivity)

• Need to consider temperature change on adsorption/desorption (isothermal vs adiabatic)

• Need to capture effect of isotherm shape on pressure profile and hence power

• Ideally, should include kinetics – however, initial goal is to permit equilibrium evaluation

www.monash.edu.au

12

Capacity for CO2 – effect of temperature

• Heat of adsorption (exothermic) leads to reduction in working capacity since adsorbent T increases on adsorption and decreases on desorption

• Instead of isotherm only, need to look at adiabat

13X Isotherm and Adiabat

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

CO2 Partial Pressure (bar)

Lo

ad

ing

(g

mo

le/k

g)

isotherm

adiabat

Isotherm at 300K, Adiabat starts at 300K

T=300K

T=370K

T=343K

www.monash.edu.au

13

Example T-swing during CO2VSA process

www.monash.edu.au

14

Development of Proposed Parameter

• Consider 1kg adsorbent with feed gas of known composition and specified pressure and temperature

• Perform adiabatic pump-down calculation tracking composition, temperature, moles removed (CO2 andN2) and energy consumed

• Derive CO2 working capacity (WCCO2), N2 working capacity (WCN2), working selectivity (WS=WCCO2/WCN2), and specific energy (ECO2=energy/mole CO2 removed)

• Repeat for a range of operating temperatures and pressure ratios

• Very easy and fast calculations

www.monash.edu.au

15

Basis for Comparison

• 3 adsorbents: NaX,

NaY, Activated

Carbon, ADSORBENT ZERO

LOADING

�H(kJ/mol)

NaX 44.4

NaY 37.0

AC 11.5

www.monash.edu.au

16

Isotherms for CO2

CO2 Adsorption Isotherms at 25C

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

0 20 40 60 80 100 120 140

P(kPa)

n(g

mo

le/k

g)

NaX

NaY

AC

www.monash.edu.au

17

Isotherms for N2

N2 Adsorption Isotherms at 25C

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.00 20.00 40.00 60.00 80.00 100.00 120.00 140.00

P(kPa)

n(g

mo

le/k

g)

NaX

NaY

AC

www.monash.edu.au

18

Comparison of Isothermal and Adiabatic Working Capacity (Yf=0.1)

• Isothermal working capacity exceeds adiabatic working capacity by several factors

• Adsorbent requirement scales directly with CO2

working capacity

• Adiabatic working capacity has a maximum unlike isothermal working capacity

CO2 Working Capacity - NaX

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

290 310 330 350 370 390

T(K)

WC

(gm

ol/kg

)

isothermal

adiabatic

www.monash.edu.au

19

Comparison of Isothermal and Adiabatic Working Selectivities (Yf=0.1)

• Isothermal working

selectivity exceeds

adiabatic working

selectivity by several

factors

• Adsorbent purity and

recovery scales with

CO2 working

selectivity

• Both show a

maximum but at

different locations

CO2 Working Selectivity - NaX

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

4.50

290 310 330 350 370 390

T(K)

WS

isothermal

adiabatic

www.monash.edu.au

20

Comparison of adsorbents – Adiabatic Working Capacity (Yf=0.15)

CO2 Working Capacity

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

290 310 330 350 370 390

T(K)

WC

(gm

ol/kg

) NaY

AC

NaX

www.monash.edu.au

21

Comparison of adsorbents – Adiabatic Working Selectivity (Yf=0.15)

Working Selectivity

0.00

2.00

4.00

6.00

8.00

10.00

12.00

290 310 330 350 370 390

T(K)

WS NaY

AC

NaX

www.monash.edu.au

22

Comparison of Energy Required

Energy Requirement

-8.50

-8.00

-7.50

-7.00

-6.50

-6.00290 310 330 350 370 390

T(K)

E(k

J/g

mo

le)

NaYAC

NaX

www.monash.edu.au

23

Ranking depends on which parameter to chose

• WC and WS suggests NaX is superior for T >

300, NaY superior for T<300

• AC shows lowest regen. energy

• Ultimate ranking should reflect capture cost

relative to a base case (NaX) (trend at least!)

• Capture Cost depends on capital (use 1/WCCO2

as a surrogate) and power (use WS as a

surrogate). Both also depend on recovery (use

WS as surrogate)

www.monash.edu.au

24

“Capture Figure of Merit” or CFM

• Must necessarily be approximate since relative contributions to capital and operating are site specific

• Need 2 parameters to truly reflect each situation• For a given ratio of capital : power costs, we

combine the parameters into a single CFM:

ref

* CFM; Relative CFM

CFM

WC WSCFM

E� �

www.monash.edu.au

25

How well does it perform?

• Rigorous PSA simulations conducted with 3 adsorbents

• Ranking of overall performance compared with CFM prediction

• Assume capital contributes ~50% towards overall capture cost

www.monash.edu.au

26

Performance of Adsorbents based on CFM

Performance of Adsorbents Based on CFM

over temperature range 300-400K

0

0.2

0.4

0.6

0.8

1

1.2

0 0.2 0.4 0.6 0.8 1

CFM

Dim

en

sio

nle

ss C

ap

ture

Co

st

AC

NaY

NaX

www.monash.edu.au

27

Selection of Operating Temperature based on CFM

NaX Performance Based on CFM

over temperature range 300-400K

0

1

2

3

4

5

6

7

300 320 340 360 380 400

Tfeed(K)

CF

M

NaX

www.monash.edu.au

28

Laboratory and Pilot Scale Work

• Vacuum Swing Cycle with multiple beds

• Minimum pressurization of feed gas

• Evacuation of beds to low pressure (< 10kPa)

• Energy is used on recovered stream only, not on total feed stream

• Cyclic process to ensure longevity of adsorbent

www.monash.edu.au

29

Process Development and Testing

• Experimental pilot scale

plant

provides “real” operating data validates process simulations

– 3 beds, 1m x 8 cm i.d.

• Simulation

allows prediction of the effect of operating and design variables

– in-house simulation tools

www.monash.edu.au

30

PSA Cycle Design: Nine step cycle (product purge)

II III IV V VI

FP

W

I VII

P

W

ITEM Step 1 Step 2 Step 3 Step 4 Step 5 Step 6 Step 7 Step 8 Step 9

Bed 1 I II II III IV V VI VII VII

Bed 2 VI VII VII I II II III IV V

Bed 3 III IV V VI VII VII I II II

www.monash.edu.au

31

Performance 6-step Cycle no purge

9-Step cycle with purge

Purity, % 82~83 90~95

Recovery, % 60~80 60~70

Power, kW/TPDc

4-8 8~12

Power,MJ/kg CO2

0.35 to 0.7 0.7 to 1.0

Performance Data for 6 and 9 step VSA cycles

Cycle can be optimised for specific feed gas - adsorbent combinations

www.monash.edu.au

32

Water: Not just an “impurity”

• Will our existing adsorbent (13X) tolerate high concentrations of water?

• What is the impact on process performance (purity, recovery, power) of high humidity flue gas streams

• Can we operate at elevated temperature?• If water adversely affects CO2 capture plant

performance, can we “protect” the adsorbent?• What is the impact of impurities (SOx, NOX for post

combustion, NH3, H2S for pre-combustion) on adsorbents?

www.monash.edu.au

3333

Water and Impurities - Multilayers

www.monash.edu.au

34

Current P/VSA process currently undergoing testing at a Power Plant

• Multi-step cycle with 3 beds including two pressure equalizations, heavy and light rinse, heavy purge effluent recycle

• Up to 1 TPDc

• Capable of “deep” vacuum (1-2 kPa)

• Wash tower will remove impurities and reduce water level.

www.monash.edu.au

35

Demonstration Plant in fabrication shop

www.monash.edu.au

36

Demonstration Plant

www.monash.edu.au

37

www.monash.edu.au

www.monash.edu.au

www.monash.edu.au

40

www.monash.edu.au

41

Conditions – Flue Gas & Adsorbent

NaX Sorbead WS

Manufacturer UOP BASF-Engelhard

Mass in each column(3:1),kg

25.00 9.15

Mass in each column(9:1),kg

29.96 3.66

Price, US$/kg 11.18 7.10

Water capacity, wt% sat

28 >34

Structure Crystalline

Amorphous

Property Value

CO2, % 12

O2, % 8

CO, ppm 40

NOx, ppm 170-250

SOx, ppm 60-270

H2O, % ~10

www.monash.edu.au

42

Pressure Profiles with Simple Cycle (no Purge)

0

20

40

60

80

100

120

140

0 100 200 300 400 500 600 700

Time,second

Pre

ssu

re,k

Pa.a PIT201

PIT203

PIT204

PIT205

PIT206

www.monash.edu.au

43

Preliminary Results

6-step without purge (Actual plant)

Purity: ~71%

Recovery: ~60%

6-step without purge (Simulation with void volume)

Purity: ~70% Recovery: 75%

9-step with purge (Simulation with low void volume)

Purity: 96% Recovery: 75%

Power consumption: 12kW/TPD ~ 1 MJ/kg CO2 captured

www.monash.edu.au

44

Engineering Issues

• Ash carryover from wash tower

• Water condensation in adsorption

vessels

• Corrosion & Blockage in feed blower

• Large void space dilutes product % CO2

• Partial fluidization at high gas velocity

• Main capture plant disruption

www.monash.edu.au

45

Water Condensation – reduces pumping capacity

www.monash.edu.au

46

Void Volumes

Simulations

indicates a

drop in purity

of 5-10%

www.monash.edu.au

47

Conclusions

• CO2 Capture using adsorption processes is

technically feasible but issues remain

– Scale-up is significant challenge– Advantages of no waste, very limited water use– Very simple process– Significant improvements can be made using

advanced adsorbents but must be careful of water and other impurities

• Adsorbent development must carefully

consider the relationship to the process

www.monash.edu.au

48