Supersonic Post-Combustion Inertial

CO2 Extraction System

Bench Scale Project Status Update

2014 NETL CO2 Capture Technology Meeting

Pittsburgh, PA

24 June 2015

Vladimir Balepin, Ph.D.

Anthony Castrogiovanni, Ph.D.

Andrew Robertson

Jason Tyll, Ph.D.

Project Overview

• Funding

• NETL: $ 2,999,673

• Cost Share: $ 749,918

• Total: $ 3,749,591

• Project Performance Dates

• 1 Oct 2014 - 30 Sep 2017

• Project Participants

• ATK & ACENT Laboratories

• Ohio State University

• EPRI

• NYSERDA and NYS-DED

• Project Objectives

• Demonstrate inertial CO2 extraction

system at bench scale

• Develop approaches to obtain condensed

CO2 particle size required for migration

• Demonstrate pressure recovery efficiency

of system consistent with economic goals

• Demonstrate CO2 capture efficiency

2

ICES Technology Background

• Supersonic expansion of compressed flue gas results in

CO2 desublimation (high velocity → low p & T)

• Inertial separation of solid particles instigated by turning

the supersonic flow

• CO2-rich capture stream is removed and processed

• CO2-depleted stream is diffused and sent to stack

3

Thermodynamics of ICES

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0.001

0.01

0.1

1

10

100

50 100 150 200 250 300

Ma

ch N

um

be

r

Pre

ssu

re [

ba

r]

Temperature [K]

Isentropic Expansion of 14mol% CO2 in N2 Relative to Phase Diagram of CO2

Triple Point

Partial Pressure of CO2

during Isentropic Expansion in Supersonic Nozzle(p0=2bar, T0=300K)

Region of incipient condensation

Gas Phase

Liquid Phase

Solid Phase

Post Condensation

PathStatic pressure (p), static

temperature (T) and velocity (v)

in a converging-diverging

nozzle

Low static pressure and temperature in supersonic nozzle causes CO2 to precipitate as a

solid – need to remove before diffusing back to low speed 4

Key Advantages and Challenges

Advantages Challenges

No moving parts, chemicals/additives or

consumable media

Maximization of CO2 particle size with

limited residence time

Inexpensive construction (sheet metal,

concrete)

Minimization of “slip gas” removed with

solid CO2

Small footprint (current bench scale test

article is 250kW, 3” x 24” x 96”

CO2 purity (all condensable material will

be removed with CO2)

“Cold sink” availability in solid CO2

Solid CO2 processing

Costs primarily driven by flue gas

compression

Optimization of flowpath pressure

recovery

5

Gen1a

and 1b

(swirl)

Gen2 (2D)

Gen3

(2D - long)

Principal conclusion of this effort was that CO2 particles >2.5μm are required for efficient

operation - need to control particle size generated 6

Summary of ARPA-e IMPACCT

Activity

Program Plan for Current Effort

• Year 1

• Lab-scale tests (OSU) to develop understanding of factors controlling particle size and

methods to increase

• Bench scale tests at ATK to demonstrate capture efficiency and diffusion with surrogate CO2

injection (liquid throttle of CO2 to produce controlled particle size)

• Success criteria: Demonstrate 50% capture, show path to pressure recovery required

• Year 2

• Integrate methods to increase particle size in bench scale test article

• Test with surrogate flue gas (Air + CO2 + H2O)

• Success criteria: Demonstrate migration of 80% of CO2 to 20% of duct height and path to

full scale pressure recovery

• Year 3

• Integrated bench-scale testing with capture + diffuser

• Success criteria: 75% capture with path to 90%, path to full scale pressure recovery

7

Lab-scale Testing at Ohio State

University

• Test program completed at OSU supersonic aerosol facility to gain better understanding of

nucleation process, condensation rates, and particle size behavior

Initial test results proved that under our conditions, CO2 only condenses on

solid or liquid media in the flow (i.e. heterogeneous condensation)

8

H2O homogeneous nucleation

N2(g)

CO2(g)

H2O(g)

throat

H2O freezing CO2 heterogeneous nucleation and growth

H2O droplet growth

coagulation

Temperature (K)

50 100 150 200 250 300

Pre

ssure

(P

a)

1e-1

1e+0

1e+1

1e+2

1e+3

1e+4

1e+5

1e+6

1e+7

1e+8

1e+9

liquid

vapor

solid

l

s

v

H2O hom

CO2 hom

CO2 het on H2O

CO2

H2O

M =

2.2

M =

3.3

Lab-scale Testing at Ohio State

University (continued)

• Test and analysis matrix included methods of inducing

turbulent particle collisions to promote agglomeration

• These approaches proved to be too intrusive and

resulted in local temperature increase

• Attention focused on solid CO2 injection/seeding

Combination of test data and detailed modeling led to conclusion that solid

media (e.g. CO2) seeding is most viable path to 90%+ capture by causing

flue gas CO2 to condense on particles already >2.5μm

0.0

0.2

0.4

0.6

0.8

1.0

0

1 10-6

2 10-6

3 10-6

4 10-6

5 10-6

0 50 100 150 200

Nozzle ATK

r0 = 1.25 mm

r0 = 2.00 mm

r0 = 3.00 mm

r0 = 1.25 mm, T

CO2L = 225K

r0 = 1.25 mm, LongFlat nozzle

r0 = 1.25 mm, LongFast nozzle

Mas

s fr

actio

n ra

tio, g

CO

2S/ w

co2 (

-)

Rad

ius

of C

O2

drop

let, r (

m)

Distance from throat, z (cm)

wco2

= 0.2 kg/kg

p0 = 2 atm

zinject

= -1 cm

gCO2S

/wCO2

r

Fig. 2 Mass fraction ratio of CO2 solid and radius of CO2 droplet

9

One CO2 Recirculation Approach

Solid/Liquid CO2 ProcessingSupercritical

CO2 Processing

Flue gas fromcompressor

Liquid CO2recirculation

Flue gas to the stack

Slip gas recirculation

Solid particle capture

Cyclone separator

CO2 to pipeline

Melting line

Triple point

Critical

point

Pipeline

Conditions

298K,153 barRecirculation CO2

@ 298K and 80 bar

Recirculation of liquid CO2 can achieve the

desired results of additional cooling + creation of

large particles to promote heterogeneous

nucleation capable of migration. Requirement to

inject very close to throat to mitigate evaporation.

10

Flow splitter plate

CO2-enriched capture flow

CO2- depleted

flow to diffuser

Current Bench Scale Test

Arrangement (250kW)

Plenum

Expansion

duct

Turning

duct

Capture

duct

Diffuser

liquid throttle solid CO2 injectors

Location of CO2

measurement probe

(NDIR and GC) 11

Laser Images of CO2 in Flow

Upper wall

Bottom wall

Air flow only

CO2 relatively uniform across

duct

Highest concentration of CO2

entering capture duct

Diffuser

Capture duct

CO2 depleted flow

CO2 enriched flow

12

CO2 Capture Data

Last year - goal of capture >50% of

CO2 achieved for short duration tests.

Cumulative measurement error due to

GC and NDIR sensor contamination

after first several seconds

This year – gas sample approach reworked to mitigate several sources

of error including time lag, pump oil contamination + added in-situ

calibration. Preliminary review of results indicate >50% capture of

solid CO2 in several recent tests – data still in detailed review

Gas samples taken from primary flow stream were processed with on-line gas

chromatograph (GC) and NDIR sensors to access CO2 mole fraction.

13

Full Scale Pressure Recovery Predictions

Current scale limits pressure recovery

performance due to thick boundary layer relative

to duct size. We have shown path to target

pressure recovery of 40% through:

• CFD benchmarking using subscale test

results and predictions of full scale

performance

• Definition of flowpath updates required to

improve performance from 31% to 40%

overall pressure recovery

10X Photo Scale

12 Units

CFD

Expansion

End

Curve

End

3” Duct 15” Duct

Component CFD Current

Configuration

Desired

Performance

Expansion

Duct

79 % 85%

Turning Duct 88% 95%

Diffuser Duct 45% 50%

Total 31% 40% 14

• A preliminary Techno-economic assessment by WorleyParsons (WP) was carried out in 2013

• Key efficiency/economic numbers are provided in the table below:

Resulting lower COE increase for ICES technology is based on lower capital and O&M costs and improvements in the overall plant efficiency

A path to the DOE research goal of 35% COE increase is being developed based on a more detailed capex/labor model and reduced flue gas compression (PR=2.0 vs 2.5 used in WP analysis)

Metric Case 11 Case 12, Amine

Plant ICES Plant

CO2 capture no yes yes

Net plant efficiency (HHV basis) 39.3% 28.4% 34.5%

COE % increase base 77% 42%

Parasitic Load 5.5% 20.5% 7.3%

Cost per tonne of CO2 captured NA US$ 62.8 US$ 41.8

Cost per tonne of CO2 avoided NA US$ 90.7 US$ 48.4

ICES Economic Impact

15

ICES Plant Layout and Footprint

Boiler Exhaust Stack

Flue Gas

Desulfurization

(FGD)

Continuous

Emission Monitors

(CEMs)

Precipitator

Axial Compressors (3)

Air Coolers Captured CO2

processing

ICES Units

Direct Contact Cooler

(DCC)

Unique Equipment

for ICES System

ICES footprint of ~8k m2 compares to 20k to 30k m2 for an amine plant of similar capacity.

ICES nozzle and compressor stacking can further reduce footprint by 30-40%.

17

Project Schedule

MS 1. Updated BP1 PMP – complete

MS 2. Kickoff meeting - complete

MS 3. Capture duct/diffuser demonstration – complete

MS 4. Updated BP2 PMP – complete

MS1 MS2 MS3

17

MS4

Tasks Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13

Task 1. Program Management

Task 2. Lab-scale Condensation/Growth Investigation

Task 3. Analytical and Computational Investigation

Task 4. Bench-scale Capture and Diffuser Testing

Task 5. Bench-scale Condensation/Growth Testing

Task 6. Integrated System

Task 7. Plant Integration and Techno-economic Analysis

Budget Period 1 Budget Period 2 Budget Period 3

Quarters

Summary

• ICES Technology holds considerable promise as an alternative to adsorbents and

membranes

• Current NETL effort focused on solving key technical challenge of particle size

• Testing and analysis results to-date support strategy of solid CO2 recirculation

as most viable approach

• Ongoing work to optimize CO2 injection arrangement to minimize

evaporation upstream of supersonic section and to redesign turning duct to

increase pressure recovery performance

18

Acknowledgements

• NETL • Andy O’Palko

• Lynn Brickett

• ATK • Bon Calayag

• Florin Girlea • Michele Rosado

• Dr. Daniel Bivolaru

• Kirk Featherstone

• ACENT Labs • Dr. Pat Sforza

• Randy Voland

• Robert Kielb

• Ohio State University • Professor Barbara Wyslouzil

• Dr. Shinobu Tanimura

• EPRI

• Dr. Abhoyjit Bhown

• Adam Berger

• NYSERDA

• NYS-DED

19