2009 DOE Merit Review

Energy Storage R&D

Thermal Management Studies and
Modeling
Ahmad A. Pesaran, Ph.D.
National Renewable Energy Laboratory
Golden, Colorado

DOE Hydrogen Program and Vehicle Technologies

Annual Merit Review
Crystal City, Virginia
May 18- 22, 2009
es_12_pesaran

This presentation does not contain any proprietary or confidential information.

1
 2009 DOE Merit Review

NREL Energy Storage Program

Our projects support the three major elements of DOE’s integrated
Energy Storage Program to develop advanced energy storage
systems for vehicle applications.

• Battery Development, Testing, Analysis

– Thermal characterization and analysis Will be discussed
in this
– Energy storage simulation and analysis presentation.
• Applied Battery Research
– Li-ion thermal abuse reaction modeling
• Exploratory Battery Research Will be discussed by
Anne Dillon on
– Nano-structured metal oxide anodes Thursday morning.

2
 2009 DOE Merit Review

Outline
Discussion of three activities funded

We will discuss most of

these for each section:

- Objective
1. Thermal Characterization and
- Barriers Analysis
- Approach
- Accomplishments 2. Energy Storage Simulation
-
-
Future Work/Plans
Summary
and Analysis
- Response to comments 3. Li-ion Thermal Abuse Reaction
- Publications
Modeling

3
 Overview
2009 DOE Merit Review

Timeline Barriers
• Project start date: Oct 2004 • Decreased life at high
• Project end date: Sep 2013 temperatures (15 years target)
• Percent complete: 60% • Safety concerns due to thermal
runaway
• High cost due to high cells cost
Partners and system integration
• USABC
• A123 Systems Budget
• CPI/LG Chem • Total project funding
• EnerDel ─ DOE share: $5.3M
• Johnson Control Saft ─ NREL & Industry: $1.3M
• General Motors • Funding received in
• General Atomics
─ FY08: $1.20M
• NASA ─ FY09: $1.40M
4
1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Objectives/Milestone/Approach
• Objectives (Task 6 of the DOE’s Vehicle Technologies R&D Plan)
– Measure thermal properties of batteries and ultracapacitors
– Model thermal performance of batteries
– Support USABC and FreedomCAR developers
• Milestones
– Thermal evaluation of advanced batteries (August 2008 and June
2009)
– Electro-chemical-thermal based battery models (July 2008 and August
2009)
• Approach
– Work with developers on thermal characterization, evaluation, and
modeling of cells, modules, and packs
– Use NREL’s collective experimental and modeling capabilities to
support developers in addressing issues of battery thermal
management and performance
5
1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Thermal Characterization Approach

Cells, Modules and Packs

Tools: Test Profiles: Measurements:

• Calorimeters • Normal operation • Heat capacity
• Thermal imaging • Aggressive operation • Heat generation
• Electrical cyclers • Driving cycles • Efficiency
• Environmental • US06 • Thermal performance
• UDDS
chambers • HWY
• Spatial temperature distribution
• Cell-to-cell temp. imbalance
• Dynamometer • Discharge/charge rates • Cooling system effectiveness
• Vehicle simulation • Constant current
• Geometric charge/discharge cycles
z tools • FreedomCAR profiles
• Thermal analysis tools
Results reported to DOE, USABC, and developers

6
 1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Thermal Characterization:
Johnson Controls- Saft Low-Temp. HEV Cells
Calorimetry Thermal Imaging at 12C Rate
• Heat capacity & heat generation & efficiency • Temperatures: Ambient
• Temperatures: -30 to +30˚C • Profiles: 100% SOC to 0% SOC
• Profiles: USABC 25 & 50 Wh cycles, CC discharge

+30°C
Heat Efficiency

-30°C Heat Efficiency > 95% at 30°C

Constant Current Discharge (A)

7
 1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Thermal Characterization:
Johnson Controls- Saft PHEV VL22M Cells
Calorimetry
• Heat capacity & heat generation & efficiency Thermal Imaging
• Temperatures: -30 to +30˚C • Temperatures: Ambient
• Profiles: CC discharge • Profiles: 100 Amp Geometric Cycle, 5C Discharge

+30°C

Heat Generation
Heat Efficiency

Constant Current Discharge (A)

Heat Efficiency > 90% for currents < 5C rate at 30oC
8
1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Thermal Characterization:
Johnson Controls-Saft PHEV VL41M Cells
Thermal Imaging
• Temperatures: Ambient
Terminals
• Profiles: CD PHEV Cycle, CS PHEV Cycle,
Geometric Cycles, CC Discharge

Body Temp
60 CS Cycles
5 CD Cycles

Center Cell Temp Profile

CD: Charge Depleting
CS: Charge Sustaining

Calorimetry Future Work

• Heat capacity & heat generation & efficiency
• Temperatures: -30 to +30˚C
• Profiles: CC discharge, CD PHEV, & CS PHEV

9
1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Electrical Characterization:
Lithium Ion Capacitor Cells
• C/1, 10C, 100C, and HPPC Testing

Energy: 14 Wh/kg HPPC Discharge/Regen Power

Power: 1500 W/kg

Power

2200 F cell

This asymmetric capacitor had high resistance;

10 the next generation is claimed to be better.
 1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Thermal Characterization:
Lithium Ion Capacitor 2200 F Cells
Calorimetry Thermal Imaging
• Heat capacity & heat generation & efficiency • Temperatures: Ambient
• Temperatures: +30˚C • Profiles: 50C, 100C, and Geometric Cycle
• Profiles: CC discharge cycles

Discharge - Exothermic
Heat Generation

Increasing Discharge Current

Charge - Endothermic
Increasing Charge Current

Calorimeter Response to Constant Current Charge/Discharge

11
1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Electrical Evaluation:
CPI HEV Module
• Electrical Study – HPPC and Voltage Performance under US06
• Consisting of eight (8) G4.3 LG Chem MnO2 cells.

HPPC Impedance

OCV vs DOD
Average Cell Discharge Impedance

Average Cell Charge Impedance

12
1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Thermal Evaluation:
CPI HEV Module
• Tested simulating real conditions and operation US06 Test Profile
• Used different power profiles and ambient conditions
• Excellent thermal performance (2°C ΔT)
Continuous US06 Cycling

Slow Fan Speed High Fan Speed

Temperature

13 Thermal performance improved with higher air flow rate

1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Thermal Evaluation:
Nesscap Ultracap Module
• Tested as part of USABC deliverable
• Eighteen (18) symmetric carbon-carbon ultracapacitors
• Tested under realistic conditions and operation
• Used different power profiles and chamber temperatures
Heat from cells are conducted through
the ends to the case and the rejected
through the top metal heat sink/fins. Balancing Board Affecting Terminal Temps

Terminal Temperatures: One Hour Geometric Cycle

14
1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Thermal Evaluation:
Nesscap Ultracap Module
• Continuous US06 cycling for two hours (next to
• Balancing board did a good job equalizing cells balancing
board)
• Energy drain for balancing could be a concern

Balance Board Temp

Max to Min Voltage Difference

Temperature difference less than 1.5°C
except for Cell #1 which heated due to
balancing board.
15
1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Completed Fabrication of A New Calorimeter for

Large, Liquid-Cooled HEV & PHEV Modules
• Used to measure heat generated
from large batteries under real
driving profiles and conditions
– Liquid cooled capability
• The new calorimeter can test
batteries 6 times larger than the
existing NREL calorimeter
• Could be used for other automotive
components such as power
electronics & motors.
Completed System with Heating/Cooling Unit

Test Chamber in Isothermal Bath

16 Flux Gauges of Test Chamber Test Chamber
1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Large Calorimeter Calibration and Battery Testing

20.40 Joule electrical input released in a resistor
Measured response = 20.94 Joules

A large module in
20 Joules = 0.00478 calories Test Chamber
One gram of fat is equivalent to
6 calories

Existing Calorimeter

Large Calorimeter
Error better than 3%

Large calorimeter’s results compares well with

17 the results of our existing calorimeter
1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Battery Thermal Modeling at NREL

Cell Characteristics

Design Process

Module Cooling 3D Operating

System Conditions
Strategy Component
Analysis
Analysis

Battery Thermal Responses

18
1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Multi-physics Battery Simulation Tool

for Better Design and Management
Background
• Wide range of time and length scales physics
• Design improvements at different scales required
• Need to better understand the interaction among different scale physics

Objectives
• Develop computer-aided design tools for better cell
design and management by working with industry
• Expand knowledge on the impacts of designs,
usages, and managements on performance, life
19 and safety of battery systems
1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Approach
Multi-Scale Multi-Dimensional (MSMD) Model
• Capture macroscopic electron and heat transports, while maintaining
model resolution to capture Li diffusion dynamics and charge transfer
kinetics in electrode level scale
• Use separate domains for 1-D Newman-type electrochemical model and
macro-scale heat and current transport model
• Physically couple the solution variables defined in each domain using
multi-scale modeling schemes

• Validate model for PHEV cell (electrical and thermal)

• Perform trade-studies for improved cell design and management
20
1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Model Validation
Thermal imaging test of three 41 Ah cells
1) Cycle: USABC PHEV10 profile
(5xCD, 60xCS)

2) 1-D EChem model well-matched to

voltage data. Critical for correct heat generation prediction.
3) Thermal-only model used to quantify
boundary conditions on center cell.

Thermal
Image Model

4) 3-D EChem/Thermal model gives good

prediction of cell skin temperature rise.

5) Future: Validate cell-internal temperatures. Internal ΔT=1.4°C

(for relatively benign USABC cycle)
21
1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

AABC 08, Tampa, May 2008

Impact of Tab Location & Size

In a Large Prismatic Cell

Accomplishments
• Micro-scale electrochemical processes and
macroscopic heat and electron transports
closely interact.
• Severe spatial non-uniformity can be caused by
poorly designed macroscopic design features.
22
1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

214th ECS, Honolulu, Oct 2008

Impact of “Aspect Ratio”

In a Large Cylindrical Cell

Accomplishments
Poorly designed electron and heat transport
pathways can cause excessive nonuniform use
of materials which lead to deterioration of
performance and shorten the life of the battery.
23
1. Thermal Characterization and Analysis Activity
2009 DOE Merit Review

Future/Planned Work
• Continue working with HEV and PHEV battery developers on
thermal characterization and analysis of batteries
• EnerDel
• A123 Systems
• CPI/LG Chem
• Johnson Controls - Saft
• Others
• Use large calorimeter to measure heat from large PHEV modules
and sub-packs
• Validate and refine the thermal-electrochemical model with
experimental data and use it for developer’s batteries
• Demonstrate the application of computer-aided battery design tool
for PHEV prismatic cells
• Investigate cost effective approaches for thermal control of batteries
when a PHEV is parked in hot environments

24
 2. Energy Storage Simulation and Analysis
2009 DOE Merit Review

PHEV Battery Performance/Life/Cost Trade-off Analysis

Objectives
• Optimize energy storage
system designs to:
• minimize cost,
• meet performance requirements,
• meet life requirements
• ensure reliability,
• accelerate PHEV market
penetration & fuel displacement.
• Evaluate real-world scenarios
• climate, driving cycles, charging
frequency.
• Life model represents greatest uncertainty (significant focus for FY09)
• complex dependency on t1/2, t, # cycles, T, V, ΔDOD.

• Life model requirements

• use accelerated and real-time calendar and cycle life data as inputs,
• is mathematically consistent with all empirical data,
• is extendable to arbitrary usage scenarios (i.e., it is predictive).
25
 2. Energy Storage Simulation and Analysis
2009 DOE Merit Review

Life Modeling Approach*

NCA dataset fit with empirical, yet physically justifiable formulas Li-ion NCA chemistry

Calendar fade Cycling fade

Resistance Growth (mΩ)

• SEI growth (partially • active material structure
suppressed by cycling) degradation and
• Loss of cyclable lithium mechanical fracture
• a1(∆DOD,T,V) • a2(∆DOD,T,V)

Resistance
Growth R = a1 t1/2 + a2 N

Relative
Capacity Q = min ( QLi , Qactive )

QLi = d0 + d1 x (a1 t1/2) Qactive = e0 + e1 x (a2 N)

Data: John C. Hall, IECEC, 2006.

*K. Smith, T. Markel, A. Pesaran, “PHEV Battery Trade-off Study and Standby Thermal Control,” 26th
26 International Battery Seminar & Exhibit, Fort Lauderdale, FL, March, 2009.
 2. Energy Storage Simulation and Analysis
2009 DOE Merit Review

Example Trade-off Studies

Impact of requirements on battery size: Impact of climate on power fade
Useable ΔDOD and cost Calendar fade model with Typical Meteorological Year
PHEV10: Assumed the battery has to last 10 years at (TMY) climate dataset
various temperatures Assumed battery temperature = ambient
Higher P/E increases useable ΔDOD

Pay for
energy Pay for
power
Lowest
cost

Some Li-ion technology must be sized with

Reducing temperature exposure from 35°C to significant excess power to last 15 years.
25° reduces PHEV10 battery cost by $1000.
27
 2. Energy Storage Simulation and Analysis
2009 DOE Merit Review

Accomplishments
• Developed empirical life model for Li-ion carbon/NCA chemistry.
(Additional comparison with DOE ANL/INL Gen II and NASA JPL datasets ongoing)
• Quantified impacts of life requirements (years and Temp.) on battery size and cost.
• Quantified calendar degradation for various climates in the USA.
Milestones
• Plug-In battery design - trade-off analysis (May 2009)
• Initial evaluation of EV battery swap concept (September 2009)
Future Work/Plans
• Extend models to understand
– Implications of real world scenarios (climate,
driving profiles, charging frequency, …)
– Impact of various Li-ion chemistries.
• Work with others to obtain PHEV field data to
validate the life model.
• Investigate the impact of ambient temperature
and battery life on various EV infrastructure
approaches such as
– Fast charge
28 – Battery swap
3. Li-ion Thermal Abuse & Internal Short Modeling
2009 DOE Merit Review

Applied Battery Research for Transportation

High Energy Battery Technology
Task 3: Abuse Tolerance Studies
Task 3.1: Abuse Behavior Modeling and Diagnostics

Multiple-Physics Safety Modeling

with Emphasis on Internal Short
• Safety is a major barrier for Li-ion batteries
• Need to develop safe and abuse-tolerant designs
• We are developing models in support of this
Modeling for Understanding Impacts of
Battery Design Parameters on
Thermal Runaway in Lithium-Ion Cells/Modules
29
3. Li-ion Thermal Abuse & Internal Short Modeling
2009 DOE Merit Review

FY09 Objective – Model for Internal Short

Develop and improve the “chemical reaction” model to evaluate
recommended designs and/or materials that could enhance the safety
tolerance of lithium-ion batteries, with emphasis on internal shorts
Research Focus – Understanding Multi-physics of Internal Short
• Understanding electrochemical response for short
• Understanding heat release for short event
• Understanding function and response of safety designs
Milestones
• Enhance 3-D Li-Ion battery abuse model (July 2009)
• Validate “electro-chemical-thermal” based battery abuse model (Sep 2009)
Approach – Development & Validation of Multi-physics Model
• Perform multiple physics modeling to expand understanding of
internal shorts by linking the electrochemical cell model to the
electro-thermal-abuse reaction kinetics model
• Collaborate with Sandia National Lab to plan and perform
30 experimental tests for model validation
3. Li-ion Thermal Abuse & Internal Short Modeling
2009 DOE Merit Review

Accomplishment 1
NREL’s multi-physics model combination demonstrates that heating
pattern at short events depends on various physical parameters such as
nature of short, cell size, rate capability.
30 mΩ Short

4s 8s 12s

7 Ω Short

9 Schematic shows the concept of how we combine the

electrochemical short cell model and 3D exothermic kinetics
model.
9 Contours show the difference in heating (temperature) for
different electrical resistance shorts at the same cell. 3m 6m 9m
31
3. Li-ion Thermal Abuse & Internal Short Modeling
2009 DOE Merit Review

Accomplishment 2
1D electrochemical short cell model results imply that detecting electric
signal of internal shorts during battery operation is not easy for large
format cells.
2.5 Ah cell
10sec 4C discharge 10sec 1C charge cycle, [I] [V]
Rs = 100 mΩ

System current
Cell current
Short current

25 Ah cell
[I] [V]
9 Figures compare electrical
responses (voltage and current) of
short cells having different
capacities under repeated
discharge and charge cycle.
32
3. Li-ion Thermal Abuse & Internal Short Modeling
2009 DOE Merit Review

Accomplishment 3
Electrical, thermal and electrochemical natures significantly change for
different type of internal shorts.

9 Electric potential distribution under short between metal (Al, Cu) foils (e.g.,
metal debris penetration through electrode & separator layers)

Planned Work
• Perform analysis for evaluating recommended safety designs
such as functional separators (ceramic coated, shut-down
feature) for various cell design parameters (materials, electrode
thickness, cell capacity, etc)
• Design experimental apparatus for model validation through the
33 collaboration with Sandia National Laboratory
3. Li-ion Thermal Abuse & Internal Short Modeling
2009 DOE Merit Review

3D Thermal Propagation Model in a Module

• Developed a 3D cell and module geometry capturing impact of cell-
to-cell interconnects on cell-to-cell thermal propagation.

CAD drawing of a 10-cell module Grid for the 10-cell module

• 10 large cylindrical cells connected in series were inserted into a

insulation holder
• Heat conduction through electrical connector dominates heat
transfer between the cells in this module design
34
3. Li-ion Thermal Abuse & Internal Short Modeling
2009 DOE Merit Review

Thermal Propagation in a Module

after Thermal Runaway of One Cell

Perspective View
Top View 10 minutes apart between each frame/image Between frames

35
 2009 DOE Merit Review

Overall Summary
• NREL collaborates with industry and other national labs as part of
the DOE integrated Energy Storage Program to develop advanced
batteries for vehicle applications.
• We moved toward achieving our goals, accomplish technical
objectives, and delivered our milestones in the areas of
1. Thermal characterization and analysis
2. Energy storage simulation and analysis
3. Li-ion thermal abuse reaction modeling
• Our activities support DOE goals, FreedomCAR targets, the
USABC Tech Team, and battery developers.
• We developed tools and supported industry either through one-on-
one collaborations or dissemination of information in international
conferences and journals.

www.nrel.gov/vehiclesandfuels/energystorage/publications.html

36
 2009 DOE Merit Review

Acknowledgements
• Programmatic Support from DOE’s Vehicle Technologies Program
– David Howell
• Contributions by NREL Colleagues
– Matt Keyser
– Gi-Heon Kim
– Kandler Smith
– John Powell
– Tony Markel
• Technical Guidance and Exchange with USABC and Energy Storage
Tech Team
– GM
– Chrysler
– Ford
– Southern California Edison
• Input and Prototypes from Battery Developers
– A123 Systems
– Saft
– Johnson Controls
– CPI
– LG Chem
– JSR Micro
37 – NessCap