Fuel Cell System Modeling and Control for Vehicular ...mypages.iit.edu/~chmielewski/presentations/seminar/...- Smart Grid Operation - Water Resource Management - Hybrid Vehicles Market

Department of Chemical and Biological Engineering

Illinois Institute of Technology

Fuel Cell System Modeling and Control

for Vehicular Applications

Donald J. Chmielewski Associate Professor

Center for Electrochemical Science and Engineering Department of Chemical and Biological Engineering

Illinois Institute of Technology Chicago, IL



City of Chicago and IIT

Northwestern

University

University

of Chicago

Illinois

Institute of

Technology



City of Chicago and IIT

Northwestern

University

University

of Chicago

Illinois

Institute of

Technology



Academics at IIT

Engineering

Architecture

Psychology

Science and Letters

Business

Law

Industrial Design



Architecture at IIT

Mies van der Rohe



Architecture at IIT

Mies van der Rohe

Helmut Jahn



Armour College of Engineering

Biomedical Engineering

Chemical & Biological Engineering

Civil, Architectural & Environmental Engineering

Electrical & Computer Engineering

Mechanical, Materials & Aerospace Engineering



Engineering Alumni

Marvin Camras Martin Cooper Paul Galvin



Engineering Alumni

Marvin Camras

Donald Othmer

Martin Cooper

Kenneth Bischoff

Paul Galvin



Engineering Alumni

Marvin Camras

Donald Othmer

Martin Cooper

Kenneth Bischoff

Bernard Baker

Paul Galvin



Chemical & Biological Engineering

Energy & Sustainability

- Fuel Cells & Batteries

- Fluidization and Gasification

- Hybrid Systems

Advanced Materials

- Interfacial Phenomena & Colloids

- Transport Phenomena in Complex Fluids

- Biomaterials

- Fuel Cell Materials

- Nanotechnology

Biological Engineering

- Multiscale Modeling of Proteins

- Biosensors & Hydrogels

- Diabetes Modeling & Technology

- Pharmaceutical Engineering

Systems Engineering

- Complex Systems Analysis

- Advanced Process Control

- Process Monitoring and Diagnosis



Fuel Cells and Batteries

Jai Prakash

- Electro-catalysis: material synthesis

and characterization

Vijay Ramani

- Hybrid materials for PEMFC: hydration and degradation

Satish Parulekar

- Modeling of SOFC electrodes

Donald Chmielewski

- Modeling, design and control of fuel

cell systems



Process Systems Engineering (Chmielewski Lab)

Control Theory

Profit Control

- Chemical Processes

- Inventory Planning

- Smart Grid Operation

- Water Resource Management

- Hybrid Vehicles

Market Responsive Control

- Power Plant Dispatch

- Building HVAC with Thermal Energy Storage

Energy Systems

Power Systems

- Dry Gasification Oxy-Combustion (DGOC) Process

- Control of Oxygen Enhanced Boilers

- Oxygen as Energy Carrier

Fuel Cell Systems

- SOFC

- Fuel Processors

- PEMFC

- Hybrid Vehicles



Technical Outline

• Start-up of a On-board Fuel Processor

• PEMFC Hydration Dynamics and Control

• Control and System Design for Hybrid Vehicles



Fuel Processor System at Argonne

Water

WG

1

AirWater

Fuel

AirW

G2

WG

3

WG

4

PrO

x1

PrO

x2

PrO

x3

ATR

Water

WG

1

AirWater

Fuel

AirW

G2

WG

3

WG

4

PrO

x1

PrO

x2

PrO

x3

ATR



Fuel Processing Reactors

PEMFCPreferential

Oxidation

(PrOx)

Water-

Gas

Shift

(WGS)

Reformer

Hydrocarbon Feed

Large Hydrocarbons Cracked:

Low H2 to CO ratio Most CO converted to CO2: ~ 1% CO remaining

CO levels down to ~ 10 ppm



Fuel Processing Reactors

PEMFCPreferential

Oxidation

(PrOx)

Water-

Gas

Shift

(WGS)

Reformer

Hydrocarbon Feed

Large Hydrocarbons Cracked:

Low H2 to CO ratio Most CO converted to CO2: ~ 1% CO remaining

CO levels down to ~ 10 ppm



ATR Reactor

Vaporized gasoline,

Steam

Liquid water

Heat exchangerAir (25 °C)

Hot air

Nozzle

7 m

m1

2 m

m1

2 m

m

96 mm

Catalyst bed

Heater rod

Thermocouple1 2 3 4

5 6 7

8 9 10

Metal wall

thickness=1.7 mm

High Space Velocity

(GHSV ~ 50,000/h)

Noble Metal Catalyst

(Rh on a Gd-CeO2 substrate).

Operating Temperature

~ 700 – 1000o C



Reactor Model (Axially Dependent, Nonlinear Dynamic Version)

)()()(,

)()(

0s

jg

jg

jccc

gj

g

kAx

m

N

i

iijj

g

j

s

j

g

jc rMk1

)()()(

,0

)()()()(

)()( ˆ0 sggccc

gg

pg TThA

x

Tcm

)()()(

)()( )(ˆ wsw

w

ww

pw TTxh

t

TSc

Mass Balances:

Catalyst Phase:

Gas Phase:

Energy Balances:

Gas Phase:

n

1i

c

)()(

llreactor wa fer toHeat trans

)()()(

,

)()()(

...)(1ˆ

ii

sg

cc

sw

ww

s

axe

ss

p

s

rHTTh

TTxhx

T

xt

Tc

Solid Phase:

Reactor Wall:



Reactions within ATR

211 Oykr

222)1(

2

m

m

Ca

OHC

yk

yykr

Oxidation: OHnmCOOnmHC nm 222 2/)4/(

Steam Reforming: 22 )2/( HnmmCOOmHHC nm

Water Gas Shift: 222 HCOOHCO

)/(22233 eHCOOHCO Kyyyykr

Papadias, et.al. (2006), Ind. Eng. Chem. Res.

Strongly exothermic

Strongly endothermic

Combined



Reactor Start-up: A 2 Step Procedure

Partial Oxidation Mode (to quickly increase temperature)

ATR Mode (for greater CO conversion)

Hydrocarbon Fuel Air

ATR

Reactor

Steam

Hydrocarbon Fuel

Air

PO

Reactor



Steady-State Axial Profiles

0.00

0.05

0.10

0.15

0.20

0.25

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Dimensionless x-axis (x/L)

Mo

lar f

ra

cti

on

s w

et

(-)

H2

CO

H2O

CO2

Fuel

CPOX Mode: ATR Mode:

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Dimensionless x-axis (x/L)

Mo

lar f

ra

cti

on

s w

et

(-)

H2

CO

H2O

CO2

FuelO2



Model Validation

0

100

200

300

400

500

600

700

800

20 40 60 80 100 120 140 160 180 200

Time (s)

Te

mp

era

ture

(°C

)

7 mm

19 mm

Inlet temperature

30 mm

Papadias, et.al. (2006), Ind. Eng. Chem. Res.



Feedback Control



Load Change with PI Controller

0 50 100 150 2000

10

20

30

40

50

Fuel F

low

, g/m

in

0 50 100 150 200140

190

240

290

340

Time, s

Ste

am

Flo

w,

g/m

inFuel

Steam

0 50 100 150 200500

550

600

650

700

750

800

Tem

pera

ture

, oC

0 50 100 150 20070

110

150

190

Time, s

Air F

low

, g/m

inT2, measured

Air

0 50 100 150 2002

3

4

5

6

7

8

Time, s

Hydro

gen F

low

, m

ol/m

in o

r C

O,

mol%

H2

CO



Feed-forward plus Feedback Control



Load Change with FF Controller

0 50 100 150 2000

10

20

30

40

50

Fuel F

low

, g/m

in

0 50 100 150 200140

190

240

290

340

Time, sS

team

Flo

w,

g/m

in

Fuel

Steam

0 50 100 150 200660

680

700

720

740

760

780

Tem

pera

ture

, oC

0 50 100 150 20070

150

230

Time, s

Air F

low

, g/m

inT2, measured

Air

0 50 100 150 2002

3

4

5

6

7

8

Time, s

Hydro

gen F

low

, m

ol/m

in o

r C

O,

mol%

H2

CO

Solid: C7.3H14.28

Dotted: C8H18



Conclusions from Classic Control

• Feedback Control

– Good performance for small load changes.

– Poor performance for large load changes.

• Feed-forward Control

– Good performance for large load changes

– Model mis-match a major concern



Nonlinear Model Predictive Control

Steady State Optimizer

(SSO)

Model Predictive Control

Nonlinear Dynamic Trajectory Optimizer

PI

ATR

Level 1

Level 2

Level 3

Level 4 PI PI

u

Input &Output

Reference

)(ku



Challenges to Model Predictive Control

- Optimization based

- Small sample intervals for feedback

- Accurate model for feed-forward action



Applying

Model Reduction (Galerkin Approximation)

N

j

j

s

j

s ztxtzT1

)()( )()(),(

N

j

j

w

j

w ztxtzT1

)()( )()(),(

)(ˆ tqBAxx

Tw

N

wws

N

ss xxxxxxx ],;,[ )()(

2

)(

1

)()(

2

)(

1

We arrive at the finite dimensional model



Solution method CFD ROM1

Run-time, s 110 30

Prediction Horizon is 180s

Matlab with Core 2 Duo CPU 2.33G Hz and 1G DDR RAM

Comparison of Computational Effort



Reduced Mass Balance Model

nmHCw

ra ~

122

11

22 2OHHC wwKrnm

za

erzr 2

22 )(

zaerzr 2

33 )(

23 )5.0~

~

(

2

2 rnmwM

wMr

nm

nm

HCH

HHC

zaerzr 1

11 )(

222)4//()(

,1 OO

g

Omc Mnmwkr

m

kAa

g

Omcc

)(

,

12



Solution method CFD ROM1 ROM2

Run-time, s 110 30 0.04

Prediction Horizon is 180s

Matlab with Core 2 Duo CPU 2.33G Hz and 1G DDR RAM

Comparison of Computational Effort



ATR Model Comparison

50 100 150 2000

200

400

600

800

1000

Time,s

Solid

Tem

pera

ture

,oC

Reduced

CFD

@7 mm



Nonlinear Model Predictive Control

Steady State Optimizer

(SSO)

Model Predictive Control

Nonlinear Dynamic Trajectory Optimizer

PI

ATR

Level 1

Level 2

Level 3

Level 4 PI PI

u

Input &Output

Reference

)(ku



Inlet Temperature Disturbance

C)20( inT

60 65 70 75 80 850

50

100

150

200

250

300

Time, s

Flo

w r

ate

, g/m

in o

r dm

3/m

in

Inputs

Water

Air

Fuel

60 65 70 75 80 85650

700

750

800

850

900

950

Time, s

Tem

pera

ture

, oC

Outputs

NMPC

Open-Loop

Feedforward



Feedstock Model Mis-Match

60 65 70 75 80 850

50

100

150

200

250

300

Time, s

Flo

w r

ate

, g/m

in o

r dm

3/m

in

Inputs

Water

Air

Fuel

(C7.3H14.28 C8H18)

60 65 70 75 80 85600

700

800

900

1000

Time, s

Tem

pera

ture

, oC

Outputs

NMPC

Open-Loop

Feed-forward



Technical Outline






Dynamic Model of PEMFC

Parameters based

on 1 kW scale.

Humidified

hydrogen feed

Air cooling is

assumed.

Cooling

Air In

Jacket

Exhaust

Anode

In

(H2, H2O)

Ecell

H2

Cathode

In

(air)

Cathode

Exhaust

O2

H2O

N2

Solid Material Current Collector

Insulator

H2O

Anode

Exhaust

Polymer Membrane

Catalyst LayersGas Diffusion

Layers (GDLs)



Dynamic Model of PEMFC

2

2 2 2

2

2 2 2

2 2

2

2 2 2

2

2 2

,

( )

( ) ( ) ( )

,

( )

,

( )

( ) (

,

( )

H in

an an H in an H H mem

an

H O in an an an

an an H O in an H O H O mem

in an

an an H H O mem

O in

ca ca O in ca O O mem

ca

H O in ca ca

ca ca H O in ca H O

dCV F C F C r A

dt

dCV F C F C J A

dt

F C F C r J A

dCV F C F C r A

dt

dCV F C F C

dt

2

2 2

) ( )

( )

ca

H O mem

in ca

ca ca O H O mem

J A

F C F C r J A

( )

( )

( )

( )

in inca

ca ca ca ca ca sol ca

p ca

in inan

an an an an an sol an

p an

jac in in

jac jac jac jac jac sol jac

p jac

sol

p sol ca solcasol

dT UAV F T F T T T

dt C

dT UAV F T F T T T

dt C

dT UAV F T F T T T

dt C

dTC V UA T T

dt

( ) ( )jac sol an sol gen memjac anUA T T UA T T Q A

Material Balances Energy Balances



Electrochemical Model

mtohmactnercell EEEEE

OH

OHsoloner

P

PP

F

RTEE

2

22

2/1

ln2

osol

act jjF

RTE /ln

2

1

)/(22

)( o

O

ca

O

o

oo CCjj

mem

ohm

tjIRE

jjj

F

RTE

L

Lsolmt ln

2

11

)(

22 ca

OmtL CKFj

GDL

ca

GDLmt tDK )(



Hydration Model for MEA

MEA

Anode

In

(H2, H

2O) H

2

Cathode

Air in

Cathode

Exhaust

O2

H2O

N2

Solid Material Current Collector

H+

H+

H+

H+

H+

H+

H+

H+

Anode

Exhaust

H2O

2

2

( )

( )

mem

H Omem

H O diff drag

C jJ J J D

z F

Boundary Conditions

2

)(2)(

22

z

CD

t

C mem

OH

e

mem

OH

mOH

ca

OH

mem

OH

e

an

OH

mem

OH

e

zrJF

j

z

CD

zJF

j

z

CD

at0

0at0

22

2

2

2

)(

)(

)(

)(



Water Transport in the Membrane

ELECTRO-OSMOTIC DRAG

DIFFUSION



Concentration Profiles

GDL Membrane GDLAnode Gas Cathode Gas

δm δ cδ a

2

( )an

H OC

( )ˆ mem

oC

( )

20mem

H OC

2

( ) ( )mem

H OC z

( )ˆm

memC2

( )ca

H OC

2

( ) ( )mem

H O mC



Power/Temperature Control

PEMFC

Fjac Tsol

+-

PITsol

(sp)

Power

Controller

Pe(sp)

Ecell

jPe ,

,c a

o oF F




100 200 300 400 500 600

0.2

0.25

0.3

0.35

0.4

0.45

0.5

Po

we

r D

en

sity (

wa

tts/c

m2)

Time (seconds)

Pe

Pe

(sp)







100 200 300 400 500 600

4

6

8

10

12

14 ! W

ate

r C

on

ten

t

Time (seconds)

(0)

(mem

/2)

(mem

)




0 100 200 300 400

0.05

0.1

0.15

0.2

Po

we

r D

en

sity (

wa

tts/c

m2)

Time (seconds)

Pe

Pe

(sp)







0 100 200 300 4004

4.5

5

5.5

6

6.5

7 ! W

ate

r C

on

ten

t

Time (seconds)

(0)

(mem

/2)

(mem

)



Manipulation of Hydration Profile

0 100 200 300 4004

4.5

5

5.5

6

6.5

7

! W

ate

r C

on

ten

t

Time (seconds)

(0)

(mem

/2)

(mem

)

PEMFC

+- Gc

Tsol(sp)

Power/

Temp

Controller

Pe(sp)

Ecell,Fjac

jPe ,

,c a

o oF F

2

( )mem

H OCTsol

u m

2

( ),mem sp

H OC



Anode Bubbler Temperature (Open Loop Test)

0 100 200 300 400 500 6005

6

7

8

9

10

11 !

Wa

ter

Co

nte

nt

81 86o oC C



Solid Temperature Set-Point (Open-Loop Tests)

80 85o oC C

0 100 200 300 400 500 6003

4

5

6

7

8

! W

ate

r C

on

ten

t



Combined Approach

0 100 200 300 400 500 6007

8

9

10

11

12

13

14 -

Wate

r C

onte

nt



Manipulation of Hydration Profile

PEMFC

+- Gc

Tsol(sp)

Power/

Temp

Controller

Pe(sp)

Ecell,Fjac

jPe ,

,c a

o oF F

2

( )mem

H OCTsol

u m

2

( ),mem sp

H OC 0 100 200 300 400 500 6007

8

9

10

11

12

13

14

- W

ate

r C

onte

nt



Most Frequent Question

100 200 300 400 500 600

0.2

0.25

0.3

0.35

0.4

0.45

0.5

Po

we

r D

en

sity (

wa

tts/c

m2)

Time (seconds)

Pe

Pe

(sp)



Most Frequent Question

Transportation

Applications??

100 200 300 400 500 600

0.2

0.25

0.3

0.35

0.4

0.45

0.5

Po

we

r D

en

sity (

wa

tts/c

m2)

Time (seconds)

Pe

Pe

(sp)



Technical Outline






Hybrid System with DC-DC Converters

FC kfc

iafcifc

Vfc

iab

ib

Vb

Rb

Eb

ia

Ra

La

Vakb



DC-DC Converter Model

FC kfc Va

iafcifc

Vfc

DC-DC converter relations:

Fuel cell V-I relation (polarization curve):

Combined V-I relation:

fcfcafcfcfca kiiVkV /

)( fcfc ifV

)( afcfcfca ikfkV

fcfcfcafcafcfcfcfcfc RkkVEiRiEV / then ,iscurveonpolarizati If



The Open-Loop Process

Vehicle

kbatvveh

kfc



Vehicle Speed Control

Vehicle

kbatvveh

PI

kfc+

-Vfc

(sp)x

Vfc

min

max

Vfc(sp)

PI

+-

vveh(sp)



Vehicle Speed Control Simulation

2 4 6 8 10 12 14

0

2000

4000

6000

8000

10000

12000

Power Profles

time, sec

2 4 6 8 10 12 14

30

40

50

60

70

80

90

100

110

120

130

Voltage

time, sec

Armature

Fuel Cell

Battery

Battery

Fuel Cell

0 5 10 150

200

400

600Motor Speed

0 5 10 150

5

10

15

20Vehicle Speed

time, sec

motor

[rad/s]

Vspeed

[mph]

V(sp)



Lesson from Reactor Control

CSTR

Valve

position

FjacketPI+

-

Fjacket(sp)

T

PI+-

T (sp)

CA

PI

CA(sp)

+-




CSTR

Valve

position

CA

PI

CA(sp)

+-




Vehicle

kbat

vveh

PI+-

vveh(sp)



Power Load Control

Vehicle

kbatvveh

+-

kfc

x

Pfc

+-

x

Pbat

Pfc(sp)

Pbat(sp)

VfcFUEL CELL

VOLTAGE

CONTROLLER

Vfc(sp)

PI

PI



Vehicle

kbatvveh

+-

kfc

x

Pfc

+-

x

Pbat

Pload(sp)

Pbat(sp)

VfcFUEL CELL

VOLTAGE

CONTROLLER

Vfc(sp)

PI

PI

Hybrid Power Load Control



Separation of Time-Scales

FC Kfc

iafcifc

Vfc

iab

ib

Vb

Rb

Eb

ia

Ra

La

VaKb

5 10 15 20 25 30 35

-200

0

200

400

600

800

1000

Power Profles [W]

time, sec

Pload

(sp)

Battery

Fuel Cell

Armature

Vehicle

kbatPmot

+-

kfc

x

Pfc

+-

x

Pbat

Pmot(sp)

Pbat(sp)

VfcFUEL CELL

VOLTAGE

CONTROLLER

Vfc(sp)

PI

PI



Vehicle Speed Control

Vehicle

kbatvveh

+-

kfc

x

Pfc

+-

x

Pbat

Pload(sp)

Pbat(sp)

+-

x

vveh(sp) Vfc

FUEL CELL

VOLTAGE

CONTROLLER

Vfc(sp)

PI

PI

PI



Speed Control Simulation

0 10 20-2000

0

2000

4000

6000

8000

10000

12000

14000Power Profles [W]

time, sec

0 10 200

20

40

60

80

100

120

140

time, sec

Voltage [V]

Armature

Pload

(sp)

Battery

Fuel Cell

Battery

Armature

Fuel Cell

0 5 10 15 200

200

400

600Motor Speed

0 5 10 15 200

5

10

15

20Vehicle Speed

time, sec

[rad/sec]

Vehicle Speed [mph]

Vehicle Speed Request



Hybrid Fuel Cell Vehicle (Double Storage Configuration)

iascapiscap

Rscap

Escap

iarm

Rarm

Larm

DC-DC

Converter

iabatibat

Rbat

Ebat

DC-DC

Converter

iafcifc

EfcDC-DC

Converter

Fuel

Cell

Power Bus

warm

Earm

kfc kbat kscap



Supervisory Control

Vehicle

Power

System

+ -Pscap

(sp) Pscap

kscap

Supervisory

Controller

Pmotor

+ -Pbat

(sp) Pbat

kbat

+ -Pfc

(sp) Pfc

kfc

PI

PI

PI



Disturbance Modeling

Vehicle

Power

System

+ -Pscap

(sp) Pscap

kscap

High

Level

Controller

Pmot(sp)

+ -Pbat

(sp) Pbat

kbat

+ -Pfc

(sp) Pfc

kfc

PI

PI

PI



Drive Cycle Characterization

0 200 400 600 800 1000 1200 14000

20

40

60

time (sec)

Sp

eed

(m

ph

)

0 200 400 600 800 1000 1200 1400-40

-20

0

20

40

time (sec)

Po

wer

to M

oto

r (k

W)



Hybrid System Optimization

such that:

- Operation within constraints limits

- Motor power demands met

- Supervisory controller embedded

ˆ ˆ ˆmin{ }fc fc bat bat sc scc m c m c m



Separation of Time Scales

20 20.02 20.04 20.06 20.08 20.1-15

-10

-5

0

5

10

15SuperCap Power, kW

time(hr)

20 20.2 20.4 20.6 20.8 21-2

-1.5

-1

-0.5

0

0.5

1

1.5

2Battery Power, kW

time(hr)

20 25 30 35 40 451.9

1.95

2

2.05

2.1Fuel Cell Power(kW)

time(hr)



Acknowledgements

• Students: Yongyou Hu (ATR) Kevin Lauzze (PEMFC) Syed K. Amed (PEMFC and Hybrid Vehicle) • Collaborators: Shabbir Ahmed and Dennis Papadias (ANL, ATR) Herek Clack (MMAE-IIT, ATR) Said Al-Hallaj (UIC, Hybrid Control) Ali Emadi (ECE-IIT, Hybrid Control) • Funding: IIT Graduate College and Armour College of Engineering Argonne National Laboratory National Science Foundation (CBET – 0967906)

Fuel Cell System Modeling and Control for Vehicular ...mypages.iit.edu/~chmielewski/presentations/seminar/...- Smart Grid Operation - Water Resource Management - Hybrid Vehicles Market

Documents