Fuel Cells Systems Analysis

R. K. Ahluwalia, X. Wang and R. Kumar

2011 DOE Hydrogen Program Review

Washington, DC

May 9-13, 2011

This presentation does not contain any proprietary, confidential, or otherwise restricted information.

Project ID: FC017

2

Overview

Timeline Start date: Oct 2003 End date: Open Percent complete: NA

BarriersB. CostC. PerformanceE. System Thermal and Water

ManagementF. Air ManagementJ. Startup and Shut-down Time,

Energy/Transient Operation

Budget FY11 funding: $650K

DOE share: 100% FY10 funding: $650K

Partners/Interactions Honeywell CEM+TWM projects DTI, TIAX 3M, Gore, NJIT ISO-TC192 WG12, HNEI,

JARI, LANL IEA Annexes 22 and 25 FreedomCAR fuel cell tech team

This project addresses system, stack and air management targets for efficiency, power density, specific power, transient response time, cold start-up time, start up and shut down energy

3

Objectives and Relevance

Develop a validated system model and use it to assess design-point, part-load and dynamic performance of automotive and stationary fuel cell systems. Support DOE in setting technical targets and directing

component development Establish metrics for gauging progress of R&D projects Provide data and specifications to DOE projects on

high-volume manufacturing cost estimation

4

Approach

Develop, document & make available versatile system design and analysis tools. GCtool: Stand-alone code on PC platform GCtool-Autonomie: Drive-cycle analysis of hybrid fuel

cell systems

Validate the models against data obtained in laboratory and at Argonne’s Fuel Cell Test Facility. Collaborate with external organizations

Apply models to issues of current interest. Work with FreedomCAR Technical Teams Work with DOE contractors as requested by DOE

5

Collaborations

– Argonne develops the fuel cell system configuration, determines performance, identifies and sizes components, and provides this information to TIAX for high-volume manufacturing cost estimation

– Establishing closer ties with DTI, conducting joint life-cycle cost studies

Air Management Honeywell Turbo TechnologiesStack 3M, NuveraWater Management Honeywell Aerospace, Gore, NJITThermal Management Honeywell Thermal SystemsFuel Economy ANL (Autonomie)H2 Impurities JARI, LANL, ISO-TC-192 WGSystem Cost DTI, TIAXDissemination IEA Annex 22 and 25

6

Summary: Technical Accomplishments

System analysis to update the status of technology Stack: Determined the performance of NSTF stacks at low

temperatures and on drive cycles Air Management: Evaluated the dynamic performance of

Honeywell’s compressor-expander-motor (CEM) and compressor-expander-motor/generator (CEMG) modules

Fuel Management: Evaluated the dynamic performance of parallel ejector-pump hybrids

Water Management: Analyzed the dynamic performance of planar and supported liquid membrane (SLM) humidifiers

Thermal Management: Analyzed the dynamic performance of microchannel automotive radiators and PEFC stack during cold start on drive cycles

Drive Cycle Simulations: GCtool-Autonomie simulations for fuel economy, ownership cost, and optimum FCS operating parameters

Cost: Collaborated with DTI in projecting system cost for different sizes and efficiencies and estimating the life cycle costs

7

Argonne 2011 FCS Configuration

2011 FCSMEA- 3M NSTFC MEA- 20-µm 3M membrane- 0.05(a)/0.1(c) mg/cm2 Pt- Metal bipolar plates

Air Management System- Honeywell CEMM- Air-cooled motor/AFB

Water Management System- Cathode MH with precooler

Thermal Management System- Advanced 40-fpi

microchannel fins

Fuel Management System- Parallel ejector-pump hybrid

S1 – Pressurized FCS, 2.5 atm stack inlet pressure at rated power S2 – Low-pressure FCS, 1.5 atm stack inlet pressure at rated power Dynamic performance of the components and the system

8

Cost vs. Performance Trade-off Study Saving in Pt by accepting rated-power efficiency <50% (system S2):

45% at η=45%, 59% at η=40%, 62% at η=35% Cost estimates from DTI correlations with Argonne data for

components and subsystems Projected saving in stack cost by accepting rated-power efficiency

<50%: 34% at η=45%, 46% at η=40%, 50% at η=35% Projected saving in system cost by accepting rated-power efficiency

<50%: 19% at η=45%, 25% at η=40%, 26% at η=35%

10

20

30

40

50

60

35 40 45 50System Efficiency (%)

Cos

t ($/

kWe)

FCS

BOP

Stack

Cost Breakdown (45% Efficiency S2)

2.46

2.59

5.70

4.81

7.84 23.02

PEFC Stack

Air Management

Fuel Management

Water Management

Thermal Management

Balance of System

9

Cell Voltage to Reach 60% Peak Efficiency Parasitic losses: 7.1% of net power at peak efficiency point 37% CEM M/C efficiency near idle H2 losses to crossover & purge Target not met with improved M/C

efficiency aloneCurrent Cell Stack CEM M/C Peak FCS

Pressure Temperature Density Voltage Efficiency Efficiency Efficiency(atm) (oC) (A/cm2) (mV) (%) (%) (%)1.2 75 0.1 784 60.2 37 561.2 75 0.1 784 60.2 60 57.4

Cell V to reach 60% peak efficiency 32 mV ∆V at 0.1 A/cm2, 1.1 atm,

SR = 2 Needed cell V lower at 0.2 A/cm2

Required cell V higher at higher P Required cell V even higher at

SR = 5750

775

800

825

850

1.10 1.15 1.20 1.25 1.30 1.35Stack Inlet Pressure, atm

Cel

l Vol

tage

, mV

0.2 A/cm2

H2 Cross-over: 3 mA/cm2

Cathode SR: 2CEM M/C Efficiency: 60%

0.1 A/cm2

20

30

40

50

60

0 10 20 30 40 50 60 70 80FCS Net Power, kW

FCS

Effic

ienc

y, %

50%

45%40%

35%

Rated Power Efficiency

S2 Configuration1.5-atm peak pressure0.15 mg-Pt/cm2

20-mm membrane

10

Catalyst Activity to Reach Target Cell Voltages Reference activity is for ternary Pt68(CoMn)32 catalyst with

0.1 mg/cm2 Pt loading– Reaching 60% peak efficiency is very difficult if the cell has to

operate at >2 SR or >1.25 atm at low current densities – Easier to reach the peak efficiency target at 0.1 A/cm2 than at

0.2 A/cm2 for P < 1.3 atm, SR = 2– Inset for 4X absolute activity which can be reached with higher

Pt loading, or higher mass activity, or combination of the two

0

10

20

30

40

50

60

70

80

90

100

1.10 1.15 1.20 1.25 1.30 1.35Stack Inlet Pressure, atm

Act

ivity

Mul

tiplie

r

0.1 A/cm2

0.2 A/cm2

H2 Cross-over: 3 mA/cm2

Cathode SR: 5CEM M/C Efficiency: 60%

0

2

4

6

8

10

1.10 1.15 1.20 1.25 1.30 1.35Stack Inlet Pressure, atm

Act

ivity

Mul

tiplie

r

0.2 A/cm2

0.1 A/cm2

H2 Cross-over: 3 mA/cm2

Cathode SR: 2CEM M/C Efficiency: 60%

4X Activity Map

11

Air Management System Determined the performance of Honeywell’s CEMM in S1 and S2

systems for two modes of cooling the motor– GCtool model, performance maps from component data

Expander reduces CEM parasitic power by 4 kW in S1, 1.5 kW in S2– Recovering cooling air reduces CEM power by 0.4-1.6 kW

Turndown is a function of the minimum rpm and may be limited by the surge line– Turndown >10 and minimum rpm <35k desirable else parasitic

power at Idle >500 W

5

7

9

11

13

15

1.5 2.0 2.5Stack Inlet Pressure, atm

CEM

Par

asiti

c Po

wer

, kW

Tamb=40°C

with Expander

w/o Expander

b

a

b

a

(a) (b)

0

100

200

300

400

500

600

700

10 20 30 40 50

Turndown

CEM

Par

asiti

c Po

wer

, W

30 krpm

35 krpm

40 krpm

Surge Line

12

Dynamic Performance of Integrated CEM Module Dynamic simulations on hybrid UDDSand HWFET cyclesOperation as motor/generator (CEMG) Faster response during deceleration SR >2 in deceleration, >>5 at low

power Parasitic power >> steady-state

values during acceleration, lower or even negative during deceleration

-6

-4

-2

0

2

4

6

8

10

0 20 40 60 80FCS Power, kW

CEM

Par

asiti

c Po

wer

, kW

Generator Mode

Motor Mode

0

10

20

30

40

50

60

0 20 40 60 80FCS Power, kW

O2 U

tiliz

atio

n, %

CEMG Mode

0

10

20

30

40

50

60

0 20 40 60 80FCS Power, kW

O2 U

tiliz

atio

n, %

CEM Mode

Solid lines indicate results from steady-state simulations

13

Fuel Management System Parallel ejector-pump hybrid

– Ejector for >45% power (zone I), blower for <30% power (zone II), hybrid for intermediate power (zone III)

– Motive gas pressure regulated to <8 atm Dynamic simulations with single-speed blower (always on)

– Ejector pumping power up to 400 W (parasitic reduced by 1 kWe)– H2 utilization maintained around 50% in zone I, << 50% in zone II

0

20

40

60

80

100

0 20 40 60 80

FCS Power, kW

H2 U

tiliz

atio

n (%

)

0

100

200

300

400

500

Ejec

tor P

umpi

ng P

ower

, W

H2 Utilization

EjectorPower

H2 feed rate proportional to pressure differential between cathode and anode– >60% H2 utilization during

depressurization– H2 utilization affected by

impurity buildup that reduces ejector entrainment

14

Purge Losses Dynamic simulations on UDDS and HWFET cycles, hybrid mode ISO-TC 192 specs for N2 (100 ppm) and He (200 ppm) impurities

– Purge schedule determined by N2 buildup due to crossover from air– Cyclic buildup of reactive hydrocarbon impurities– NH3 does not accumulate due to significant crossover to cathode– Cumulative degradation due to H2S– Anode purged (15 sl) when inerts (N2 + He) build up to 10 mol%– H2 loss and purge schedule depend on duty cycles and allowable

inert concentration (maximum)

0

2

4

6

8

10

0 500 1000 1500 2000Time, s

Mol

e Fr

actio

n, %

UDDS HWFET

N2

He

0

10

20

30

40

0 10 20 30 40 50 60FCS Power, kW

Purg

e Lo

ss, %

100

150

200

250

300

Sche

dule

, s

UDDSHWFET

Schedule

Purge Loss

UDDS

HWFET

15

Water Management System (Humidifier) Completed analysis of Honeywell data for full-scale, half-scale, and

1/10th sub-scale membrane humidifiers – Data received for planar humidifier with a composite membrane

Summary of model results– Optimum dry-air inlet temperature for maximum flux (last year) – Thinner membrane: higher water flux, but lower optimum T and

mechanical support may be needed– Lower P: Higher flux, but required water transfer rate also greater – Cathode RH reported at stack inlet T (air concurrent with coolant)– Anode RH reported at stack outlet T (air countercurrent with fuel)

40

60

80

100

120

0 20 40 60 80FCS Power, kW

Cat

hode

Inle

t RH

, %

0

20

40

60

80

100

0 20 40 60 80FCS Power, kW

Ano

de In

let R

H, %

Anode humidification is due to recycle (no anode humidifier in the system)

16

Supported Liquid Membrane (SLM) Moisture diffusion coefficient higher in liquid membranes than in solid

membranes Composite SLM by Zhang, J. Membrane Science 276 (2006) 91-100

– LiCl solution immobilized in hydrophilic cellulose acetate (CA) membrane, 50-70 µm thick, 0.22 µm pores

– Hydrophobic PVDF membranes as protective layers, 45 µm thick, 0.15 µm pores

– Composite SLM considered experimental, not optimal

0

1

2

3

4

5

6

45 55 65 75 85 95Dry Air Inlet Temperature, oC

Wat

er F

lux,

g m

-2s-1

P = 1.5 atmWet Air T = 75oCWet Air RH = 100%20 µm Membrane

Nafion

SLM

15˚C Approach TDP

10˚C Approach TDP

Modeled Performance For same membrane thickness

higher flux at high T Flux limited by resistance in the

protective layers Many SLM configurations

reported in literature– Higher flux – Eliminate pre-cooler

17

Thermal Management System – Heat Loads Heat loads at rated power and on 6.5%-grade at 55 mph

– Stack heat load proportional to Stacked HT radiator, LT radiator and AC condenser, 40-fpi MC fins

– On grade, stack T allowed to rise to 95oC, P to 2.3-2.4 atm– Expander needed even in S2 – Depending on vehicle platform and FCS rated power, 35-40%

efficiency at rated power may be acceptable

PEFCPEFCPEFCP ηη /)1( −

500

600

700

mV

80 kW

61.5 kW

30

40

50

60

% 80 kW

61.5 kW

507090

110130

35 40 45 50System Efficiency, %

kW 80 kW

61.5 kW

Cell Voltage

Stack Efficiency

Heat Load 200

400

600

800

35 40 45 50System Efficiency, %

W

Radiator Fan

Coolant Pump

12

14

16

18

m2

Pumping Power

Fin Area

18

Dynamics of Cold Start Improving water transport model using 3M’s polarization and water

balance data at 30-80oC and data without anode MPL Cold-start simulation on 1 UDDS and 1 HWFET starting at 27oC

– Flooding limits stack power to 32 kW at 30oC, 50 kW at 45oC– Stack heat-up time and temperature depend on the drive cycle– FCS efficiency is a function of stack temperature and drive cycle

0.5

0.6

0.7

0.8

0.9

0.0 0.2 0.4 0.6 0.8 1.0 1.2Current Density, A cm-2

Cel

l Vol

tage

, V

27°C

75°C

35°C

0

20

40

60

80

100

0 500 1000 1500 2000Time, s

FCS

Pow

er, k

WC

oola

nt O

utle

t T,

o CCoolant T

FCS Power

UDDS HWFET

0

20

40

60

80

100

0 20 40 60 80FCS Power, kW

FCS

Effic

ienc

y, % CEMG Mode

0

20

40

60

80

100

0 20 40 60 80FCS Power, kW

FCS

Effic

ienc

y, % CEM Mode

19

Optimum FCS Control Parameters Drive cycle simulations for 35-50% rated-power efficiency and 65-120

kW FCS rated power– Mid-size hybrid vehicle, Kansas City drive cycles– FCS provides the hotel loads, 250-W idle power

Determined optimum control parameters: power threshold for turning FCS on, threshold for idling FCS, minimum FCS power, battery power

Minimum FCS power and traction power (at motor) for turning on the FCS is a function of the rated power and the rated-power efficiency– Depending on battery SOC, FCS operates at low traction power

Preliminary results, control strategy will depend on stack/battery durability considerations

0

5

10

15

20

60 70 80 90 100 110 120FCS Power, kW

Thre

shol

d Po

wer

, kW

FCS Minimum

FCS Off

FCS On35% FCS Efficiency at Rated Power

0

5

10

15

20

60 70 80 90 100 110 120

FCS Power, kW

Thre

shol

d Po

wer

, kW

FCS Minimum

FCS Off

FCS On45% FCS Efficiency at Rated Power

20

Summary: System Performance on Drive Cycles Relationship between rated-power efficiency, FCS rated power, FCS

cost (DTI input), fuel economy, ownership cost– Optimum FCS control parameters

Net present value (NPV) shows that 85-kW FCS with 40% rated-power efficiency offers the best solution (3.70/5 $/gge gasoline/H2)– NPV: present value of investments and future savings, compared to

reference ICEV Fuel economy higher with the largest FCS, small dependence on rated

power efficiency

0

500

1000

1500

2000

2500

3000

3500

4000

60 70 80 90 100 110 120FCS Rated Power, kW

Net

Pre

sent

Val

ue, $

FCS Efficiency at Rated Power: 50%

45%

40%

35%

40

45

50

55

60

65

70

75

80

65 70 75 80 85 90 95 100 105 110 115 120

FCS Rated Power, kW

Fuel

Eco

nom

y, m

pgge

50% 45% 40% 35%FCS Efficiency at Rated Power

21

Future Work Support DOE/FreedomCAR development effort at system,

component, and phenomenological levels Continue collaboration with 3M to validate, calibrate and document

the stack model– Alternate membranes, catalyst structures, and system

configurations Continue cooperation with partners to validate air, fuel, thermal, and

water management models– Establish closer collaborations with the OEMs

Support DTI and TIAX in high-volume manufacturing cost projections, collaborate in life-cycle cost studies

Collaborate with 3M to develop durability models for NSTFC electrode structures– System optimization for cost, performance, and durability– Drive cycle simulations for durability enhancement

GCtool model of PEFC systems for fork-lift applications Performance of PEFC systems for stationary applications

22

Supplemental Slides

27

Effect of Stoichiometry and Operating Pressure At low current densities, cells are sometimes operated at higher

SR because of limited CEM turndown or catalyst flooding issues– Cell voltage for 60% peak efficiency is ~50 mV higher if the

cathode SR is 5 instead of 2 at 0.2 A/cm2, 1.2 atm – Needed cell voltage is ~17 mV higher if the pressure is

increased from 1.2 to 1.3 atm at 0.2 A/cm2, SR = 2

775

800

825

850

875

900

925

950

1.1 1.2 1.3 1.4 1.5Stack Inlet Pressure, atm

Cel

l Vol

tage

, mV

0.2 A/cm2

0.1 A/cm2

0.1 A/cm2

Cathode SR = 5 0.2 A/cm2

Cathode SR = 2

H2 Cross-over: 3 mA/cm2

CEM M/C Efficiency: 60%

28

Advanced Binary vs. Ternary Catalyst

0.10

0.15

0.20

0.25

0.30

0.00 0.05 0.10 0.15 0.20

Pt Loading, mg/cm2

Mas

s A

ctiv

ity, A

/mg-

Pt

Pt68(CoMn)32

PtM

50

75

100

125

150

175

0.00 0.05 0.10 0.15 0.20Pt Loading, mg/cm2

ECSA

, cm

2 -Pt/m

g-Pt

Pt68(CoMn)32

PtM

0.5

1.0

1.5

2.0

2.5

0.00 0.05 0.10 0.15 0.20Pt Loading, mg/cm2

Spec

ific

Act

ivity

, mA

/cm

2 -Pt

Pt68(CoMn)32

PtM

0

10

20

30

40

0.00 0.05 0.10 0.15 0.20

Pt Loading, mg/cm2

Abs

olut

e A

ctiv

ity, m

A/c

m2

Pt68(CoMn)32

PtM

For same Pt loading, PtM has 1.5-1.7 times the absolute activity of PtCoMn

29

Approach to Reach 60% Peak FCS Efficiency

Redesign the CEMM controller

– Improve M/C efficiency at low power from 35% to >60%

AND operate at low SR (<2) and low pressures (<1.2 atm)

AND increase Pt loading in PtCoMn

– 40% improvement in absolute activity by increasing Pt loading from 0.1 mg/cm2 to 0.2 mg/cm2

OR use more active binary PtM catalyst (still under development, mass transfer issues at high current densities, durability yet to be demonstrated)

– Compared to 0.1-mg-Pt/cm2 ternary catalyst, the activity of binary catalyst is 70% higher for same Pt loading and 120% higher at 0.15 mg-Pt/cm2 loading

30

Water Management System Completed analysis of Honeywell data for full-scale, half-scale, and

1/10th sub-scale membrane humidifiers – Data received for planar humidifier with a composite membrane

Summary of model results– Optimum dry-air inlet temperature for maximum flux (last year) – Thinner membrane: higher water flux, but lower optimum T and

mechanical support may be needed– Lower P: Higher flux, but required water transfer rate also greater

0

1

2

3

4

5

6

7

45 55 65 75 85 95

Dry Air Inlet Temperature, oC

Wat

er F

lux,

g/m

2 .s

Approach TDP, Membrane Thickness

30oC, 20 mm

10oC, 40 mm

10oC, 20 mm

P = 1.5 atmWet Air T = 75oCWet Air RH = 100%

30oC, 40 mm

0

1

2

3

4

5

6

7

45 55 65 75 85 95Dry Air Inlet Temperature, oC

Wat

er F

lux,

g/m

2 .s

Membrane Thickness = 20 mmWet Air T = 75oCWet Air RH = 100%

30oC, 2.5 atm

Approach TDP, P

10oC, 2.5 atm

10oC, 1.5 atm30oC, 1.5 atm