Tools for Designing Thermal

Management of Batteries in
Electric Drive Vehicles

Ahmad Pesaran, Ph.D.

Matt Keyser, Gi-Heon Kim, Shriram Santhanagopalan, Kandler Smith
National Renewable Energy Laboratory
Golden, Colorado
Presented at the
Large Lithium Ion Battery Technology & Application Symposia
Advanced Automotive Battery Conference
Pasadena, CA • February 4-8, 2013
NREL/PR‐5400‐57747
NREL is a national laboratory of the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, operated by the Alliance for Sustainable Energy, LLC.
 Battery Temperature in xEVs
• Lithium-ion battery (LIB) technology is expected to be the
energy storage of choice for electric drive vehicles (xEVs) in
the coming years
• Temperature has a significant impact on life, performance,
safety, and cost of LIBs
Desired
Operating Limiting power to
Temperature reduced T increase and
degradation

Dictates the size depending

Dictates power capability through on the power and energy
cold cranking fade rate
Also limits the electric driving range Kandler Smith, NREL Milestone Report, 2008

2
 Battery Thermal Management for xEVs
• Higher temperatures degrade LIBs more quickly, while low
temperatures reduce power and energy capabilities, resulting in
cost, reliability, safety, range, or drivability implications
• Therefore, battery thermal management is needed for xEVs to:
o Keep the cells in the desired temperature range
o Minimize cell-to-cell temperature variations
o Prevent the battery from going above or below acceptable limits
o Maximize useful energy from cells and pack
o Use little energy for operation
• However, a battery thermal management systems (BTMS) could:
o Increase complexity
o Add cost
o Reduce reliability
o Consume energy for operation
o …

3
 Battery Thermal Management System
• Most in the xEV battery community agree that the value that
a BTMS provides in increasing battery life and improving
performance outweighs its additional cost and complexity
• However, the BTMS needs to be designed appropriately with
the right tools
• The National Renewable Energy Laboratory has been a
leader in battery thermal analysis and characterization for
aiding industry to design improved BTMSs
• This presentation describes the tools that NREL has used and
that we believe are needed to design properly sized BTMSs

4
 Energy Balance in a Battery

dE
= E gen − E loss + E in − E out
dt
Energy Energy Energy Input Output
Accumulation Generation Loss Energy Energy
Rate Rate Rate Rate Rate

+ E in dE
E gen dt

− E loss
5
 Heat Transfer in a Battery
(Assumption: isothermal ~ very high thermal conductivity)
dTs
mCp = Heat gen − hA(Ts − Ta ) − eδA(Ts4 − Ta4 ) − QExt _ conduction
dt
Rate of Rate of Convection Radiation Conduction
Temp Internal Heat Heat Rate Heat Rate Heat Rate
Change Generation
Heat generated (Heat gen)
in a battery consists of:
dTs • Electrochemical reactions
mCp
Heat gen dt • Phase changes
• Mixing effects
• Joule heating

hA(Ts − Ta ) + eδA(Ts4 − Ta4 ) + QExt _ conduction

Ts = Battery Temp
Method of heat rejection/addition for thermal control Ta = Ambient Temp
6
 Heat Generation Rate and Specific Heat Impact
Battery Temperature Rise
16
Assuming uniform battery temperature and
the same heat transfer coefficient for three cases
14
Temperature Rise (Degree C)

12
2C Rate (4.45 W/cell) Cp=1019 J/kg/C
10 C/1Rate (1.33 W/Cell) Cp = 1019 J/kg/C
2C Rate (4.45 W/Cell) Cp = 707 J/kg/C
8
Fast discharge
6

4
Slow discharge
2

0
0 400 800 1200 1600 2000
Time (seconds)

7
 Heat Transfer in a Battery
(Non-isothermal; case and core regions)

∂T
Core region ρCp = Heat gen + ∇ ⋅ k∇T
∂t
k : thermal conductivity

Case or boundary region

∂T( n) ∂TB
− kn = h(Ts − T∞ ) + eδ (Ts − T∞ ) + ρ BCp,B H B
4 4

∂n ∂t
n = 0. L
Heat flux Convection Radiation Heat
from the from various from various accumulation
core case surfaces case surfaces in the case

8
Case + Core Example: T Distribution in a Module

55

o
TSS= 54°C
Ttop,SS = 54 C

50 Air-cooled
Air Cooled
Q = 5 W/cell
5 W/cell

Max. Cell Temperature [ oC]

o
Tt=0 = 30 C Top
o
T = 25 C
air
hh ==
18 18
W/m KW/Km
22
45 air Middle

Bottom

40

35

30
0 10 20 30 40 50 60

Time [min]

A. Pesaran, A. Vlahinos, S. Burch. Proceedings of the 14th Electric

Vehicle Symposium, December 1997

9
 What Information is Needed to Design a BTMS?
• Acceptable temperature range for cell components at all times, i.e.,
active material, binders, separators, electrolyte, etc.
• Acceptable temperature difference within cells and from cell to cell,
depending on the chemistry and management system
• Maximum and minimum temperature limits for life specifications,
performance ratings, and safety considerations
• Thermo-physical properties of cells or components (density, specific
heat, directional thermal conductivities)
• Heat generation rate under average and aggressive drive profiles and
loads for the specific electric drive
• Heat rejection rate depending on thermal management strategy
o Fluid heat transfer coefficients or sink conductance
o Cooling fluid flow rate and sink temperature
• Configurations and dimensions of cells and proposed BTMS
• Parasitic power needed to push fluids/cooling through BTMS
10
 Tools for Designing BTMS
• Experimental Tools
o Isothermal calorimeters and battery testers
o Infrared thermal imaging
o Thermal conductivity meters
o Heat transfer characterization setup
o Battery thermal testing loop

• Modeling Tools
o First-order/lumped capacitance thermal and fluid models
o 1-D and 2-D thermal and fluid-flow performance models
o 1-D vehicle integrated thermal-flow models
o 3-D electro-thermal models
o 3-D electrochemical-thermal model
o Computer-aided engineering software

11
 Isothermal Battery Calorimeters
• We use a single-ended (one test chamber) conduction
calorimeter to measure specific heat and heat generation at
various current rates, temperatures, and states of charge (SOCs)

Sample/Module

Battery Tester

Initially fabricated by Calorimetry Sciences Corporation; later improved by NREL.

A. Pesaran, M. Keyser, D. Russell, J. Crawford, E. Lewis. Presented at the Long Beach Battery Conference,
January 1998
12
 NREL’s First Isothermal Battery Calorimeter
• Heat flux measured between the sample and a heat sink using
heat flux gauges
• The heat sink is kept at a constant temperature with a precise
isothermal bath • Max module that could be tested: 21 cm x 20 cm x 32 cm
• Heat rate detection: 0.015 W to 100 W
• Minimum detectable heat effect: 15 J (at 25°C)
• Baseline stability: ±10 mW
• Temperature range: -30°C to 60°C (±0.001°C)
• Accuracy of better than ±3%

Calorimeter response

Heat Rate (Watts)

Total heat generation =
Area under each curve

Calorimeter Calorimeter Cavity Time (Hours)

Photo Credits: David Parson & Matt Keyser, NREL
13
 Example Heat Generation Data for CC Discharge
50.0
(from max to min allowable capacity-SOC)
45.0
22-Ah Li-Ion Cell
40.0

35.0
Heat Generation (Watts)

30.0

25.0

20.0

15.0

10.0

5.0

0.0
0.0 20.0 40.0 60.0 80.0 100.0 120.0
RMS Discharge Current (Amps)

Initial Temp = -15°C Initial Temp = 0°C Initial Temp = 30°C

14
 Specific Heat (Heat Capacity)
• Can be estimated
n
from constituents
n
of cell/module
C p,ave = ∑ (C p,i • mi ) / ∑ mi
1 1
• Can be estimated using a calorimeter by measuring heat
lost/gained (Q) from the battery going from Tinitial to Tfinal
o Heat capacity is calculated by
C p,ave = Q /(mtotal • (T initial−Tfinal ))
Cell/Module Taverage Heat Capacity
(°C) J/kg/°C
NiMH – 18 Ah 33.2 677
Li-Ion 18650 33.1 1,105
Li-Ion Pouch – 4 Ah 18 1,012
VRLA – 16.5 Ah 32 660
Ni Zn – 22 Ah 20 1,167
Thermal Characteristics of Selected EV and HEV Batteries, A. Pesaran,
M. Keyser. Presented at the 16th Annual Battery Conference; Long
Beach California, January 2001

15
 NREL’s Large Volume Battery Calorimeter
Photo Credits: Dennis Schroder & Ahmad Pesaran, NREL
• Single chamber, conduction, isothermal
• Includes several patent-pending
concepts
• Test chamber submerged
• Capability to test liquid-cooled batteries
• Safety features in case of events
• Test chamber 6 times larger than the Test Chamber
NREL module calorimeter
o 2 ft x 2 ft x 4 ft Completed System with Heating/Cooling Unit

• Heat Rate: 0.05 W to 4 kW

• Accuracy of heat meas. ±3%

Flux Gauges in Test Chamber Test Chamber in Isothermal Bath

16
 NREL’s New Isothermal Cell Calorimeter
Photo Credits: Dennis Schroder & Dirk Long, NREL
• Single chamber, conduction,
isothermal
• Test chamber submerged
under isothermal bath
• Testing chamber: 15 cm W x
10 cm L x 6 cm H
• Heat detection limit: 1 mW
and 10 J
• Initial testing shows excellent
baseline stability and an error
of less than ±1.6%
• CRADA and license agreement
signed with NETZSCH to
commercialize NREL’s battery
calorimeter design

[Link]

17
 Infrared Thermal Imaging
• Quickly finds thermal signature of the whole
cell under electrical loads
• Helps understand thermal behavior, creates
diagnostics, and improves designs
• Could be used as a validation of thermal
models
• Thermal signature depends on several factors
o Geometry, thermal conductivity of case and core,
location of terminals, design of interconnects,
Thermal image of a 6.5-Ah NiMH module from
current density, current profile, chemistry, a MY 2002 Prius under 100A CC discharge
environment
• We spray a thin layer of boron nitride on all the
surfaces of the face that needs to be imaged
• We minimize reflections from other objects by
placing the cells in a non-reflective
environment
• We usually test three cells to see the impact of
power cable connected to the two end cells
Photo Credits: Matt Keyser, NREL
18
 Examples of Battery Infrared Thermal Imaging

25°C 45°C

Photo Credits: Matt Keyser & Dirk Long, NREL

[Link]
19
 Thermal Conductivity Estimation & Measurement
• Usually case and core of a cell • The core material (electrochemically active
are considered two different part) is assumed to consist of a homogenous
regions with different thermal material with average properties for resistivity
conductivity and thermal conductivity, but with different
properties in different directions
(orthotropic xyz or rθZ)

Measurement Techniques

Can use finite kx = q *∆x /∆T

element analysis ky = q *∆y /∆T
to calculate the or
effective thermal kz = q *∆z /∆T
conductivity in kr = q *∆r /∆T
each direction Provided by Peter Ralbovsky - Netzsch Instruments
20
Measuring Thermal Conductivity of LIB Components
Flash Diffusivity Method:
• Thermal diffusivity (α) is a measure of how quickly a material can
change its temperature when heat is applied
• The temperature rise on the rear surface is measured in time using an

K (T ) = α (T ) ⋅ c p (T ) ⋅ ρ (T )
infrared detector

Netzsch LFA 447 Unit

Provided by Peter Ralbovsky, Netzsch Instruments

• Measurements have shown that generally the thermal conductivity of LIB is much
lower in-plane than cross-plane
In plane ~ 0.8 to 1.1 W/m/K Cross plane ~ 28 to 35 W/m/K Photo Credit: John Ireland, NREL
21
 Battery Thermal Testing Loops
• Measuring heat transfer coefficients or conductance

Photo Credits: Ahmad Pesaran

• Hardware in the loop thermal testing

Temp dist. in a USABC module

Photo Credits: Kandler Smith, NREL
22
 Process for Battery Thermal Modeling
Cell Characteristics
• Shape and size : Prismatic/Cylinder/Oval, etc.
• Materials/Chemistries
• Voltage/current & heat gen data
Module Cooling Strategy • Thermal/Current Paths inside a Cell
Operating Conditions
• Coolant Type: Air/Liquid • Vehicle Driving Cycles
• Direct Contact/Jacket Cooling • Control Strategy
• Serial/Parallel Cooling Design Process • Ambient Temperature
• Terminal/Side Cooling • etc.
• Module Shape/Dimensions 3D
System
• Coolant Path inside a Module Component
Analysis
• Coolant Flow Rate Analysis
• Passive with phase change
• etc.

Battery Thermal Responses

• Temperature History Cells/Module/Pack
• Temperature Distribution in a Cell
• Cell-to-Cell Temperature Imbalance in a Module
• Battery Performance Prediction
• Pressure Prop and Parasitic Power
• etc.
23
 Lumped Capacitance Thermal Model for
Vehicle Simulations
• For vehicle simulation, the thermal model needs to be linked to
the battery model for temperature dependency
• A 2-node lumped thermal model (case + homogenous core) with
simple heat convection is developed for ADVISOR vehicle
simulator

24
Example of 2-D Module Thermal Modeling
Case 1. No holes and no air flow between cells

Cell core
Air Plastic
gap case

Tmax = 53°C
Delta Tcore = 13°C
Case 2. With holes and air flow between cells

Tmax = 44°C
Delta Tcore = 9°C
Photo Credits: David Parsons, NREL
25
 Electro-Thermal Analysis Approach
• Capture details of a cell including
non-electrochemical hardware with Current Density
finite element analysis
• Estimate component resistances
using geometry and materials
• Apply voltage drop to calculate
current density in components
• Estimate resistive heating (I2R) in
each component
Temperature
• Apply electrochemical heat of
reactions in the core (active parts)
• Apply heat transfer boundary
conditions on cell exterior
• Predict temperature distribution in
the cell from current density and
related heat generation distribution
26
 Example of 3-D Electro-Thermal Modeling

Design A 16 Ah Design B Design A Design B

Terminals on Power Cell Terminals on the Thermal Response Thermal Response
same side
each side
Photo Credit: Ahmad Pesaran Under 110 A RMS load
Temp Dist Design A
• The overall resistance of Cell
Design B is less than Cell Design A
• Under the same current profile,
Cell Design B generates less heat
and thus performs better thermally
Electro-thermal Analysis of Lithium Ion Batteries, Pesaran, A.; Vlahinos, A.; Bharathan, D. Proceedings of
the 23rd International Battery Seminar, Fort Lauderdale, Florida. March 13-16, 2006.
27
 Combined 3D Electrochemical-Thermal Models
Comparison of two 40-Ah Li-ion
prismatic cell designs
2 min 5C discharge
Larger over-potential working potential working potential This cell is cycled more
promotes faster discharge
uniformly, can
reaction
therefore use less
Converging current causes
active material ($) and
higher potential drop along
has longer life.
the collectors
electrochemical electrochemical
current production current production

SOC SOC

temperature temperature
High temperature
promotes faster
electrochemical reaction
p

Higher localized reaction

Current Collector (Cu)

Current Collector (Al)

Electrode

Separator

Electrode
Negative

causes more heat generation

Positive

28
 Computer-Aided Engineering of Batteries (CAEBAT)
• U.S. Department of Energy is
Physics of Li-Ion Battery Systems in
supporting development of Different Length Scales
electrochemical-thermal models Electrode Scale Cell Scale

and software design

Charge balance and Electronic potential &
transport current distribution
Electrical network in Heat generation and
Particle Scale composite electrodes transfer Module Scale
• The objective is to shorten time Li diffusion in solid phase
Interface physics
Particle deformation &
Li transport in electrolyte
phase
Electrolyte wetting
Pressure distribution
Thermal/electrical
inter-cell
configuration

and reduce cost for design and fatigue

Atomic Scale
Structural stability
Thermal
management System Scale
Safety control
Thermodynamic properties System operating

development of battery systems, Lattice stability

Material-level kinetic barrier
conditions
Environmental
Transport properties conditions

including the design and analysis Control strategy

of BTMSs
• Other software design and analysis
tools dealing with other physics
may be incorporated in CAEBAT

Thermal-electrochemical response of a pack

Courtesy of Christian Shaffer, EC Power-CAEBAT
29
 Summary
• Battery thermal management needed for xEVs
• Battery thermal management system needs to be optimized
with right tools for lowest cost
• NREL has state-of-the art experimental and analytical tools for
analysis and design of battery thermal management systems
• Experimental tools, such as the isothermal calorimeter, are
essential for obtaining data for generating input to design
tools and eventually verifying the performance of the battery
thermal management system
• Computer-aided engineering tools for the design of battery
electrical and thermal management systems are now
accessible to automotive and battery engineers

30
 Acknowledgments
• Support provided by the DOE Vehicle Technologies Program
o Dave Howell, Hybrid and Electric Systems Team Lead
o Brian Cunningham, Energy Storage Technology Manager
• Feedback from CAEBAT Subcontract Technical Leads
o Taeyoung Han (General Motors)
o Steve Hartridge (CD-adapco)
o Christian Shaffer (EC Power)
• Support from NREL Staff Contact Information:
o John Ireland Ahmad Pesaran
[Link]@[Link]
o Dirk Long
303-275-4441
o Mark Mihalic
o Marissa Rusinek

[Link]/vehiclesandfuels/energystorage
31