1

Plate Based Fuel Processing System

David YeeCatalytica Energy Systems, Inc.

May 25 – 27, 2004

This presentation does not contain any proprietary or confidential information

2

Project Objectives

Develop new catalytic reactor designs and reactor technology forprocessing gasoline to PEM quality H2

Develop improved catalyst materials compatible with these reactor systems

Design and fabricate prototype units for each reactor at the 2 to 10kW(e) scale

Demonstrate steady state and transient performanceEvaluate rapid start up performance

3

Budget

Total FundingDOE Funding $ 8.16 millionCESI Funding $ 3.50 millionProgram Total $11.66 million

FY04 FundingDOE Funding $ 2.11 millionCESI Funding $ 0.91 millionFY04 Total $ 3.02 million

4

Technical Barriers and Targets

DOE Technical Barriers for Fuel-Flexible Fuel ProcessorsI. Fuel Processor Startup/Transient OperationJ. DurabilityL. Hydrogen Purification/Carbon Monoxide CleanupM. Fuel Processor System Integration and EfficiencyN. Cost

DOE Technical Targets for Fuel-Flexible Fuel Processors in 2010Energy efficiency 80%Power density 800 W/LSpecific power 800 W/kgCost $10/kWeCold startup time to maximum power < 1 min at –20°C (< 0.5 min at +20°C ambient)Durability 5000 hoursCO content in product stream < 10 ppm steady state (< 100 ppm transient)

5

Approach

Develop initial catalyst system compatible with plate reactor system

Measure experimental kinetics over the composition and temperature

range of interest

Assemble Reactor simulation model

Define required catalyst performance

Catalyst development

Catalyst durability testing

Model reactor performance Design/Fabricate

prototype reactor

6

Project TimelineShake-down and

testing of reformer prototype

Complete prototype reformer fabrication

Demonstrated required reforming

catalyst performance on DOE approved Benchmark Fuel I

with 10ppm S

Stable & non-pyrophoricWGS catalyst demonstrated

/ kinetics measured

Established design criteria for plate

reformer with LCF life > 100,000 cycles

Validated test rig simulating plate

reactor configuration

Start program

Program completion

2001 2002 2003 2004 2005

Demonstration of WGS and PrOx

prototype reactors

2nd and 3rd

generation reformer prototype designs

7

Fuel Processing Approach

Steam reforming

Preferential CO oxidation

H2S removal

Water gas shift

PEM quality H2

Gasoline Tier 2, 30 ppm sulfur (average) gasoline

Process heat provided by catalytic combustion of gasoline or anode purge gas (outlet ~70% H2 & 16% CO dry basis)

Absorption trapping—required level to be specified (initial target <0.1 ppm S)

20% CO at inlet with 80% conversion

1% CO at inlet with <10 ppm CO at outlet

8

Reforming Outlet

Combustion Outlet

CESI Reactor Approach

Major components based on plate-type heat exchangersCombustion catalyst

Reforming catalystOne

Reforming Cell

Combustion outletReforming outlet

Combustion In

Combustion Out

Reforming In

Reforming Out

9

First Steam Reformer Prototype Fabrication

Utilize plates from an existing heat exchanger design from a gasturbine recuperator (7.5 mil = 0.2 mm thick).Cut one-tenth sector (shaded region) to fabricate a simple prototype.

Reaction area per plate is small (5.5 by 15 cm) requiring significant number of plates to achieved desired output.

Utilized CESI coating knowledge to successfully develop a coating process.

Developed a plate welding process.

10

Plate Reactor Design

Reformate

CombustionExhaust

Combustion Inlet(opposite side of plate)

ReformingInlet

11

3 kW(e) Prototype Hardware

3 kW(e) = 32 plate pairs = 7.3 cm

12

0

300

600

900

0 10 20 30 40 50 60 70 800

12

24

36

3 kW(e) Prototype Performance

Time (seconds)

Tem

pera

ture

(° C)

F ue l

Flow

Ra t

e (g

/s )

Inlet Gas Temp

Fuel Flow Rate

Plate Outlet Measured

Plate Inlet Modeled

Steam Reformer start-up achieved within 80 secondsExperimental data validates predictive model

Experimentally measured & modeled performance

Combustion only

Full-load operation after 80 seconds

Plate Mid ModeledPlate Outlet ModeledPlate Inlet Measured

Plate Mid Measured

13

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 50 100 150 200WHSV, h-1

HC

con

vers

ion,

-

Reforming & Combustion Catalysts Kinetic Model

8)··exp(0

1C

aAC yRTEk

WdtdNr −

== 22 ··)··exp(0 COOHC

cbaAHC yyyRTEk

WdtdCr −

==

Power rate law expression fits experimental data

Experimentally determined kinetics to support modeling effort

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 100 200 300 400WHSV, h-1

Con

vers

ion

to C

1's,

-

780°C

825°C

800°C260°C

280°C

340°C360°C

14

Water Gas Shift

Modeled parameters to reduce WGS reactor volumeAll plate reactor based designsKinetics based on experimental measurements

111 or 2Number of stages

Co-currentCo-currentCo or counter current

Flow Pattern

36.1 L

90%

275°C

2.0

Base Case

19.1 LCatalytica WGS Volume

80%80% to 90%CO Abatement

250°C235°C to 295°CInlet Temperature

1.50.5 to 2.0Molar Flow Ratio (cooling/reformate)

Optimized CaseRange Studied

WGS volume reduced by 47% to 19.1 L for 50 kW(e)

15

800-hour PrOx Catalyst Durability Test

No degradation of catalyst performance after 800 h on stream

1

10

100

1000

10000

90 120 150 180 210Temperature, °C

Out

let C

O c

once

ntra

tion,

ppm

Constant WHSV

Fresh Catalyst

Aged Catalyst

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 100

CO

con

vers

ion,

O2

sele

ctiv

ity, -

600 700 800Time on stream, h

Conversion

Selectivity

16

System Performance

Current CESI’s system performance versus DOE targets

thermal stress analysis > 5,0004,0005,000hDurability

steam reformer start-up only8012060sCold start-up time

precious metal costs only212510$/kW(e)Cost

reactor components only1,400700800W/kgSpecific power

reactor components only1,650700800W/LPower density

integrated heat management calculated from PRO/II SimSci software

757880%Energy efficiency

CommentsCESI 2004

2005 target

2010 target

17

Interactions and Collaborations

Argonne National LaboratoryTed Krause – Water Gas Shift catalyst

Pacific Northwest National Laboratory Greg Whyatt – Microchannel Vaporizer

Plate FabricatorsSeveral commercial companies

National Fuel Cell Research Center, UC IrvineProfessor Scott Samuelsen – Competitive Technology and Market Assessment for the Production of a Hydrogen-Containing Stream for Use in PEM Fuel Cells

18

Reviewer’s Comments

Energy costs of starting needs to be addressedModeled several start-up scenariosEvaluated energy costs of alternative start-up heating scenarios

Sulfur management critical to all fuel processing optionsAll sulfur compounds are converted to H2S in the reformerH2S easily reduced to required level by current commercial technology

Large size of WGS suggest this should be a focusModeled alternative reactor configurations to identify performance requirements to significantly reduce WGS reactor volume

19

Future Work

Remainder of FY 2004Fabricate and test more commercial plate reactor prototype design Develop more energy efficient start-up strategiesFabricate and test PrOx plate reactor prototypeDemonstrate reforming catalyst durabilityDevelop alternative WGS reactor concepts to further reduce reactor volume

FY 2005Fabricate and test low cost & commercially viable plate reactor prototype designFabricate and test WGS reactor prototypeDemonstrate WGS durability