Lithium-Ion Battery Simulation for

Greener Ford Vehicles

October 13 , 2011
COMSOL Conference 2011
Boston, MA

Dawn Bernardi, Ph.D.,

Outline

Vehicle Electrification at Ford from Nickel/Metal-Hydride to Lithium-Ion Batteries

The Hybrid Electric Vehicle (HEV)

The Plug-in Hybrid Electric Vehicle (PHEV)

The Battery Electric Vehicle (BEV)

Li-Ion Battery Chemistry

Li-Ion Battery Modeling (HEV)

Comparison of model calculations to experimental pulse/rest behavior

Contributions to overvoltage during pulse and rest periods

Model calculations of lithium distribution

Sensitivity of voltage relaxation to particle characteristics

Sensitivity of initial overvoltage to anisotropy in solid-state Li diffusivity

Fords current HEV lineup utilizes

Nickel/Metal Hydride Battery Technology

Fords Next-Generation HEVs will use

Lithium-Ion Battery Technology

Fusion
C-max
Escape

Fords Plug-In Hybrid Electric Vehicles

will use Lithium-Ion Battery Technology
Advanced lithiumion battery provides
up to 30 miles allelectric range

C-max Energi

Fords All-Electric Vehicles

Focus
Transit Connect

Lithium-Ion Battery Chemistry

(-)

DISCHARGE

CHARGE

Li+

Li+

e
Li+

PF6

Li+

Li+

Carbon

Li+
PF6

PF6

PF6
PF6

Li+

PF6

Li+

e
Li+
e
Li+

(+)

PF6

Li+

Li+

Li+

Li+

PF6
Electrolyte

Li+
PF6

Li+

Metal oxide

Lithium-Ion Cell Model

(+)
(-)

Journal of Power Sources, 196, 412-427 (2011)

Discharge Current Pulse: Model vs. Experiment

OCV
(65% SOC)

rest

Charge Current Pulse: Model vs. Experiment

rest

OCV
(65% SOC)

10

Discharge Current Pulse: Model vs. Experiment

OCV
(65% SOC)

rest

11

Current Pulse: Calculated Overvoltage Behavior

The voltage
relaxation
period is
about 200 s.

What are the significant contributors to the overvoltage, especially during the rest period?

12

Calculated Overvoltage
Primary contributors
1.

electronic resistance of solid

active material at the positive
terminal

2.

resistance to solid-state
lithium transport in positive
active material

3.

separator resistance

13

Calculations: Solid-Phase Li Composition (at Particle Surfaces)

Throughout the Electrodes

x_avg
in LixC6_surf
Dimensionless
Lithium
Concentration

0.75

0.03 s

Negative

0.7

5s
40 s

10.2
s
14.2
s

0.65
42 s

20.2
s

60 s
0.6

>100 s

0.55

Dimensionless
Concentration
y in Lithium
Liy(NCA)O2_surf

0s

40 s

Positive
0.7

40.1
s
40.6
s

0.65

42 s
46 s
60 s

0.6
>100 s
20.2 s
10.2
s
5s

0.55

0.03
s

0s

0.003
s

0.5

0.5
0

0.2

0.4

0.6

0.8

Electrode Dimensionless Distance, L_neg

Cu foil

Separator

0.2

0.4

0.6

0.8

Electrode Dimensionless Distance, L_pos

Negative:

Reaction starts at the separator interface

Positive:

Reaction starts at the Al current-collector interface

Al foil

14

Calculations: Solid-Phase Li Composition Throughout a Particle at the

Al Current-Collector Interface in the Positive Electrode
40 s

in Liy(NCA)O2
Dimensionlessy Lithium
Concentration

40.1 s

0.7

40.6 s
42 s

0.65

46 s

60 s

0.6

100 s

20.2 s

0.55

10.2 s

5s

0.03 s
0s

0.5
0

0.2

0.4

0.6

0.8

Particle Dimensionless Radial Distance

Center of Particle

Surface of Particle

These concentration variations are responsible for the overvoltage that persists during the rest period.

15

Calculations: Solid-Phase Li Composition Throughout a Particle at

the Separator Interface in the Negative Electrode
0.8

60
0s

50

0.03 s
5s

0.7

40

10.2 s
50 s

42 s
20.2 s

0.65

30

60 s
40 s

0.6

% LiC6

Dimensionless Lithium
Concentration
x_avg in
LixC6

0.75

20
100 s

0.55

10

0.5

0
0

0.2

0.4

0.6

0.8

Particle Dimensionless Radial Distance

Center of Particle

Surface of Particle

These composition variations are not responsible for overvoltage

(because of the two-phase nature of the negative-electrode active material).

16

Calculations: 2-Dimensional Animation of Solid-Phase Li Composition

Particle Dimensionless Radial Distance

Throughout Positive and Negative Electrodes

Sep.
Negative

Cu foil

Positive

Al foil

17

Calculations: Voltage relaxation time as a function of

positive-electrode particle characteristics

18

Current Pulse:

Model vs. Experiment

(in the first 3 seconds)

Could anisotropy of DLipos explain the experimental behavior?

19

Current Pulse:

Model vs. Experiment

(in the first 3 seconds)

Fixing the diffusivity at the particle periphery would reduce overvoltage.

20

Summary and Future Directions

Weight, volume, and cost are driving the shift from

nickel/metal-hydride to lithium-ion battery technology for
automotive propulsion.

Battery models can implicate resistive factors that

reduce fuel economy.

Positive electrode: electronic resistance of active material

Positive electrode: solid-state lithium transport

Low lithium diffusivity at particle peripheries may explain

the initial steep voltage descent

When compared to behavior throughout life, models can

implicate life-limiting mechanisms.

21

Thank you for your

attention!

Back-up slides

From Nickel/Metal Hydride to

Lithium-Ion Batteries
3500

3000

Specific Power (W/kg)

2500

2000

1500

1000

NiMH High Power

500

0
0

20

40

60

80

100

120

140

160

Specific Energy (Wh/kg)

24