ISSN 2032-6653

The World Electric Vehicle Association Journal, Vol. 1, 2007

Battery Thermal Management Design Modeling

Gi-Heon Kim* and Ahmad Pesaran*

Battery thermal management is critical in achieving performance and the extended life of batteries in electric
and hybrid vehicles under real-driving conditions. Appropriate modeling for predicting thermal behavior of battery
systems in vehicles helps to make decisions for improved design and shortens the development process. For this
paper, we looked at the impact of cooling strategies with air and direct/indirect liquid cooling. The simplicity of an
air-cooling system is an advantage over a liquid-cooling system. In addition to its intrinsically lower heat transfer
coefficient, another disadvantage of air cooling is that the small heat capacity of air makes it difficult to
accomplish temperature uniformity inside a cell or between cells in a module. Liquid-cooling is more effective in
heat transfer and takes up less volume, but the added complexity and cost may outweigh the merits. The surface
heat transfer coefficient, h, and the blower power for air cooling are sensitive to the hydraulic diameter of the
cooling channel (Dh). However, because of the added thermal resistances, h evaluated at cell surface is not as
sensitive to the variation of Dh in an indirect (water/glycol jacket) cooling system. Due to the high heat transfer
coefficient at small Dh, direct liquid cooling using dielectric mineral oils may be preferred in spite of high pressure
loss in certain circumstances such as in highly transient large heat generating battery systems. In general,
air-cooling should be considered first, as the power demand increases with heavier vehicles and more aggressive
driving, water/glycol jacket cooling should be considered next. Results of computational fluid dynamics model
simulation imply that capturing the internal heat flow paths and thermal resistances inside a cell using a
sophisticated three-dimensional cell model is important for more accurate prediction of cell/battery thermal
behaviors. This paper identified analyses and approaches that engineers should consider when they design a battery
thermal management system for vehicles.

Keywords: Hybrid Electric Vehicle, HEV, Battery Model, Thermal Management System

papers did not consider the detailed electrochemistry.

1. INTRODUCTION Al-Hallaj et al. [5] presented phase change material
Temperature greatly affects the performance and life (PCM) thermal management as a potential option for
of batteries, so battery thermal control must be used in high-power applications.
electric, hybrid, and plug-in hybrid electric vehicles Fig. 1 presents the concept diagram showing the
under real-driving conditions. In recent years, battery thermal management modeling process at the
automakers and their battery suppliers have paid National Renewable Energy Laboratory (NREL). The
increased attention to battery thermal management, cell characteristics (e.g., chemistry, form-factor,
especially with regard to life cycle and related warranty dimension, and materials), operating conditions (e.g.,
costs. power load profiles from the vehicle, and ambient
The basic performance of the management system is
dictated by the thermal design of each cell and module.
Gu and Wang [1] developed a two-dimensional thermal
and electrochemical coupled lithium-ion cell model to
understand the temperature effect on electrochemistry
and vice versa. Srinivasan and Wang [2] examined the
methodology of using experimental data, instead of an
electrochemical model, to determine cell heat
generation rate. But their two-dimensional approach
was limited to a simple one-set single cell, which
cannot address various thermal conditions and
geometrical effects. Pesaran et al. [3] and Bharathan et
al. [4] have demonstrated the utility of
three-dimensional models for improving the current and
temperature distributions in batteries by including the
geometrical details in model domains. However, these Fig. 1 Working flow diagram for battery thrmal
*National Renewable Energy Laboratory, 1617 Cole Blvd., management modeling process at NREL
Golden, Colorado 80401 USA, email: gi_heon_kim@nrel.gov;

© 2007 WEVA Journal, pp. 126-133

126
 ISSN 2032-6653
The World Electric Vehicle Association Journal, Vol. 1, 2007

temperatures), module/pack cooling strategy (e.g., have been covered in separate studies [3,4]. The effects
coolant type and mass flow rates, coolant channel of using different types of coolants were examined here
design, and temperature duty cycle) are all input to the (See Table 1). We selected coolant mass flow rate ( mc )
NREL¶s battery thermal management design model. The and the hydraulic diameter of coolant channels (Dh ) as
model uses these inputs for component and system system control parameters. In this study, the cell with a
analysis to predict the thermal response of the design. 50-mm diameter, a 100-mm length, and 2 W heat
Then, the promising modifications to the design can be generation was chosen for base case cell.
evaluated to determine the optimum solution while
considering factors such as cost, volume, mass, and 2.1. Fully Developed Flow Analysis
maintenance issues. Even though the heat transfer is enhanced in a
The presented study focused on examining the viable turbulent flow regime, the required blower power
cooling strategies. In order to find a high-performance greatly increases with laminar to turbulence flow
and cost-effective cooling system, it is necessary to transition. Therefore, many heat exchanger applications
evaluate system thermal response and its sensitivity as a are designed to be operated at laminar flow regimes. If
function of controllable system parameters. This paper the channel gap is small enough compared with the cell
identifies analyses and approaches that engineers diameter, the following fully developed laminar flow
should consider when they design a battery thermal relations can be applied to the presented system.
management system for vehicles.
c f Re = 24
Nu = 5.385 (1)
2. ANALYSIS AND RESULTS
A typical parallel cell cooling system was where Re= VDh /Q, Nu= hDh /k. c f is friction coefficient.
investigated using fully developed channel relations The Nuselt number is evaluated for constant heat flux
and computational fluid dynamics. Fig. 2 presents the wall boundary conditions.
schematics of a system. Pressure loss in coolant channel Channel Pressure and Power Losses Fig. 3 (a) shows
(' P), coolant temperature change between channel inlet the channel pressure losses per unit mass flow rate
and outlet (' T1 ), and temperature difference between ( ǻ3 / m c ) as a function of coolant channel hydraulic
cell surface and coolant mean temperature (' T2 ) are diameter for different coolants. Due to the large
chosen for the system responses of interest. ' P and the difference in kinematic viscosity, ' P varies in very
coolant flow rate are critical factors determining the different ranges for each coolant fluid. ' P is directly
required pump/blower power and size. ' T1 is a proportional to fluid kinematic viscosity (Q) and coolant
parameter indicating the cell temperature uniformity mass flow rate ( m  c ). If the cell diameter is much larger
that possibly could be achieved. ' T2 is closely related than Dh , ' P becomes inversely proportional to Dh 3 .
to the heat transfer coefficient, h, and shows how much Therefore the channel pressure loss changes are very
the cell temperature would differ from the coolant sensitive to Dh when it is small, especially for the
temperature or how fast the heat would be transferred at high-kinematic viscosity fluids.
a given temperature difference. On the other hand, the m Q
ǻ3 ~ c 3
maximum cell surface temperature relative to coolant Dh
inlet temperature, ' Tmax= ' T1+ ' T2 , can be used as a
parameter for controlling the limit of cell temperature wǻ3 m Q (2)
tolerance. ~  c4
wD h Dh
Note that the cell internal temperature distribution
and the maximum temperature at cell surface depend on Flow power requirements to overcome the channel
the thermal paths and resistance distributed inside a cell. friction loss were normalized by the square of coolant
Therefore, the shapes, materials, thermal connectivity mass flow rate and compared for the different coolant
of cell components, and location of heat transfer systems in Fig. 3(b). Due to the much smaller fluid
surfaces are important for predicting cell internal density and consequently larger volumetric flow rate at
temperatures. Detailed investigations into this topic
0.5Dh Table 1 Properties of coolants typically used in
T 'T2 battery cooling systems
.
. Q Dcell Coolant
mc 'T1 Air Mineral Oil Water/Glycol
Property
Lcell
x Density U (kg/m3) 1.225 924.1 1069
P
cell surface temperature
'P Specific Heat cp (J/kg K) 1006.43 1900 3323
coolant mean temperature
pressure
Thermal Conductivity k (W/m K) 0.0242 0.13 0.3892
x

Fig. 2 Schematics of a typical parallel cell cooling Kinematic

Kinetic Viscosity
Viscosity Q 2/s)
Q (m (m2/s) 1.461e-5 5.6e-5 2.582e-6
system and system responses

© 2007 WEVA Journal, pp. 126-133

127
 ISSN 2032-6653
The World Electric Vehicle Association Journal, Vol. 1, 2007

given mass flow rate, the air-cooling system requires 10

4
10
6

(a) Air (b) Air

much higher flow power for compensating channel 3
Mineral Oil
Water/Glycol
10
5 Mineral Oil
Water/Glycol
10
friction loss than the liquid coolant systems at the given 4

Power/m c2 [W/(kgs-1)2]
10

'P/mc [kPa/(kgs-1)]
coolant mass flow rate and channel height. However, in 10
2

10
3

liquid-cooling systems, not only the coolant channel 10

2

.
1
10

friction loss but also the system manifold friction head

.
1
10

and the static pressure head make a significant 10

0

10
0

contribution to the required pumping power. 10

-1
10
-1

Temperature Differences Variations of ' T1 and ' T2 are

1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8
Dh (mm) Dh (mm)

shown for the coolant mass flow rate and the hydraulic Fig. 3 (a) Channel pressure loss per unit mass flow rate as a
diameter of coolant channel respectively in Fig. 4. To function of the coolant channel hydraulic diameter. (b)
achieve temperature uniformity through a cell, it is Flow power requirement for pressure loss normalized by
square of mass flow rate as a function of the coolant
preferred to keep coolant temperature change ( ' T1 ) in channel hydraulic diameter.
the channel as small as possible. ' T1 is inversely
proportional to coolant heat capacity flow rate. 4
Coolant Temperature Increase
8
Temperature Difference Between Coolant & Cell Surface

Air
'T1 'T2
Air
Therefore, for low flow rate cooling, a little change in (a) 3.5 Mineral Oil
Water/Glycol
(b) 7 Mineral Oil
Water/Glycol

flow rate can greatly affect the coolant temperature 3 6

change, and consequently cell temperatures (especially

Temperature (oC)

Temperature (oC)
2.5 5

when air is used for the heat transfer medium that has 2 4

small cp as compared in Fig. 4). Water/glycol is the 1.5 3

most preferred among the tested coolant materials for 1 2

0.5 1
achieving temperature uniformity of cell/pack because
0 0
of its large specific heat. 0.5 1 1.5 2 2.5
Mass-Flow Rate (g/s)
3 3.5 4 1 2 3 4
Dh (mm)
5 6 7 8

1 Fig. 4 (a) Variation of coolant temperature change

ǻ71 ~ inside a system as a function of coolant mass flow
m c c p
(3) rate. (b)Variation of temperature difference between
coolant and cell surface as a function of hydraulic
wǻ71 1 1 diameter of coolant channel.
~
wm c p m c 2
Temperature difference between coolant flow and 800
Air
cell surface, ' T2 , is linear to Dh with the slope 700 Mineral Oil
proportional to 1/k as shown in Fig. 4. The 0.7-mm Water/Glycol

jacket wall thickness and 0.05-mm air layer were 600

considered between the cell surface and water/glycol water/glycol if direct contact
500
coolant channel. Due to small thermal conductivity of
h (W/m 2K)

air, ' T2 rapidly increases with Dh in air cooling. 400

Therefore, to achieve a small ' T2 (or large heat transfer

300
coefficient, h), reducing the hydraulic diameter of the Jacket 0.7 mm, air-gap 0.05 mm
Jacket 1.2 mm, air-gap 0.05 mm
channel is critical for air cooling. On the other hand, 200 Jacket 1.2 mm, air-gap 0.20 mm
' T2 is not very sensitive to variations of Dh in the
100
water/glycol cooling system because of the relatively
large thermal conductivity. 0
1 2 3 4 5 6 7 8
Dh (mm)
1 (4)
ǻ72 ~ Dh  const Fig.5 Variation of heat transfer coefficient as a
k Function of coolant channel hydraulic diameter.
where const is 0 for direct-contact cooling in the small Dh. So, the direct liquid cooling using mineral oil
relation shown at Eq 4. shows much higher h values than the other coolants at
The ' T2 curve shown in Fig. 4 also implies the heat Dh <2~3 mm in the presented case. In spite of large
transfer coefficient, h, which is inversely proportional pressure losses due to large Q and small Dh in this
to ' T2 . h is plotted as a function of Dh in Fig. 5. High h operating region, mineral oil cooling may be preferred
lowers ' T2 to reduce cell temperature. In unsteady heat for its high-heat transfer coefficient in certain
transfer, high h means fast heat removal or addition conditions.
from small temperature displacement, limiting the peak Contours of maximum cell surface temperature
temperature of the cell. Therefore, high h smoothes out relative to coolant inlet temperature, ' Tmax (=' T1+ ' T2 )
the cell temperature oscillations under transient heat are plotted as a function of coolant mass flow rate and
generating conditions. The heat transfer coefficient hydraulic diameter of coolant channels for different
evaluated at cell surface for water/glycol jacket cooling, coolant systems in Fig. 6. The values of ' Tmax in the air
due to added thermal resistances between coolant and system are much higher compared with other fluid
cell, greatly decreases compared with the value it would systems due to its small heat capacity and thermal
be if direct-contact cooling. The reduction is greater at conductivity. Contour lines for the air system, Fig. 6(a),

© 2007 WEVA Journal, pp. 126-133

128
 ISSN 2032-6653
The World Electric Vehicle Association Journal, Vol. 1, 2007

4 4 4 12

1
7

1
2.6

1. 2
5.4

1.4

1.6
4.8
3
3.2
3.4

5.8
1.6

0.8
2.2

1. 2
(a)
3.6
(b) (c)

0.6
4.2
2

4.4
3.5 3.5 3.5

4.6
10

3.8
1.8

7.2
2.8

8.2
6.4

7.6
6.2
4

5.6

7.4
7.8
6.8
3 3 3
2.4

6.6

1.8
5.2
8
mass flow rate (g/s)

8.4

4
1.
1
1
2.5 2.5 2.5
5

1.2

1.6
6

1.4
8
7
2.6

5.4

2
4.8

1.
0. 8
3
3.2
3.4

5.8
2.2

4 3.8 .6

6
3

8.8 8.6
4.2
2

4.4
4.6

2 2 2

7. 4 7.2
2.8

8.2
6.6 6.4 1.4

7.8.6
6 .2
5.6

8
4
1. 2

1.
6.8
2.4

1.5 1.5 1 2 1.5

5. 2

1.2 1. 4 1. 6

8.4

9.2
1.4 1.6

9
5

8
7
5. 4

2.2 2
8
3. 3

5.
2

1 1 1.8
3.3.4

8
4.

2.4 1
6

2
4.42.44. 6 8.6 1.4 1.6 1.6 1.8
3.8 7.2 8. 2 8. 8 .8 2.2 2.68 1.6
4 5.6 6.2 6.4 7.477.6
.8 .69
.4 09.2 00.4.610.8 1.8 2 2.4 2. 3 1.8 2
5.2 6 6.66.8 8.4 9 9 2.6
0.5 5 7 8 9.2 110 11 11 0.5 2.2 2.4 2.8 3 3.2 3 4 0.5 1.8 2 2
22
2.2
0
1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8
Dh(mm) Dh(mm) Dh(mm)

Fig. 6 Contours of maximum cell surface temperature relative to coolant inlet temperature

are dense and mostly aligned vertically at m  c > ~ 1 g/s. Fig. 9 shows temperature distribution contours of the
This means that ' Tmax is dominated by and sensitive to cell and cooling air. In the upper frame of the figure,
Dh in this operating region. On the other hand, the radial-direction length scale is exaggerated to see the
water/glycol jacket cooling system [Fig. 6(c)] contour thermal development in the air channel. Cell surface
lines are almost horizontal at m c < ~ 2 g/s, and the line temperature constantly increases in the axial direction
 c > ~ 2 g/s. This means
density is relatively sparse at m as coolant air temperature increases. However, cell
that ' Tmax is not very sensitive to Dh , and that ' Tmax internal temperature distribution is determined by the
would not be a strict limiting design factor of the thermal paths and thermal resistances inside a cell. The
water/glycol system. The lowest value of ' Tmax appears maximum temperature of 38.4 o C appears on the cell
in the mineral oil direct-contact cooling system (Fig. axis a little bit downstream from the cell center in the
6(b)) with a small Dh and large m  c operating region. presented cell specification.
Note that large pressure loss accompanies in that Axial distribution of the airflow mean temperature,
operation region. cell surface temperature, and cell center-line
System Operation Parameter Optimization An example temperature are presented in Fig. 10. Coolant air
of confining the operation zone to given conditions is temperature change, ' T1, is computed as 1.5o C. The
shown in Fig. 7. By drawing contour lines of required maximum cell surface temperature relative to inlet air
conditions, possible operating zones can be found. The temperature, ' Tmax, is 2.9o C at the channel outlet. The
colored area shown in Fig. 7 represents the operating highest battery temperature appears in the middle of the
zone satisfying Re<2300, ' P <110 Pa, ' Tmax <4.5o C centerline of the cell and displaced from coolant inlet
and ' T1 <1.5o C in an air-cooling system. Laminar temperature by 3.4o C. Due to entrance effect, air
restriction of the Reynolds number is to avoid excessive temperature rapidly increases near the channel inlet. In
friction loss due to turbulence flow transition. Point A the entrance region, the radial profiles of temperature
is the operating point for achieving maximum h for and velocity at the cell/coolant interface have steep
given conditions. B is the lowest ' Tmax operating point, gradients representing higher heat flux and wall friction.
and C is the minimum pressure loss (' P) operating Entrance effect is more clearly seen on the axial
point. distribution of the heat transfer coefficient (h) and heat
flux shown in Fig. 11.
2.2 Computational Fluid Dynamics Analysis
Air Cooling An operating point (Dh , m  c ) = (2.2 mm,
1.33 g/s), which is close to point A in Fig. 7, the ' Tmax=3.43 o C
maximum h operating point satisfying given limiting
conditions, was selected and simulated for the ' Tmax=4.5 o C
air-cooling system using an axisymmetric
Re=2300
computational fluid dynamics (CFD) model. The
model geometry and mesh are presented in Fig. 8. B
The model includes internal cell component materials
C
and geometries. The cell core winding was treated as
' T1 =1.5 oC
a continuous material having orthotropic properties A
according to layer directions. As specified previously,
the cell is 50 mm in diameter, 100 mm in length, and ' P=110 Pa
generates heat with a rate of 2 W at its core. Inlet air ' P=37 Pa
temperature was set to 35 o C. The other surfaces,
except the channel/cell interface, were set as
thermally adiabatic boundaries. Fig. 7 Confining the operation zone and parameter
optimization to given conditions

© 2007 WEVA Journal, pp. 126-133

129
 ISSN 2032-6653
The World Electric Vehicle Association Journal, Vol. 1, 2007

Coolant Channel 39

38.5

Can Cell Core

38

Temperature (oC)
37.5

Terminal Current Collector Terminal

37

36.5

36

Air Mean Temperature

35.5
Cell Surface Temperature
Fig. 8 Model geometry and mesh representing Cell Temperature at Axis
an axisymmetric cylindrical cell 35
0 1 2 3 4 5 6 7 8 9 10
Axial Distance (cm)

Fig. 10 Axial distribution of air flow

mean temperature, cell surface
temperature and cell center-line
temperature
120
250

100

Heat Transfer Coefficient (W/m2K) 200

80

Heat Flux (W/m 2)

150
60

100
40

50
20

0 0
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
Axial Distance (cm) Axial Distance (cm)

Fig. 11 Axial distribution of heat transfer

Fig. 9 Temperature distribution contours of the coefficient and heat flux
cell and the air coolant channel (top: expanded
view near can surface and coolant channel) prediction of cell/battery thermal behaviors.
Air and Water/Glycol Cooling Comparison Cooling of
The heat transfer coefficient, h, at the cell surface has a larger cell that has a 50-mm diameter, 200-mm length,
a larger value near the channel inlet, and consequently, and generates heat with a rate of 4 W at the cell core
so does the heat flux. As the flow approaches fully was simulated for a direct air-cooling system and a
developed, the value of h converges to the constant water/glycol-jacket cooling system in order to contrast
value, 56.23 W/m2 K, which is slightly lower than the the characteristics of each system. Reducing channel
predicted value, 59.24 W/m2 K, using the relations height greatly increases heat transfer coefficient at cell
shown in Eq. 1. However, the mean value of h over the surface in a direct-contact, air-cooling system. But
heat transfer surface has a little bit higher value, 60.96 reducing the channel height is limited by the channel
W/m2 K, due to the entrance zone effect. One of the friction loss, which increases sensitively with
reasons for this discrepancy comes from the fact that decreasing Dh for a given coolant flow rate. On the
heat flux distribution is not quite constant along the other hand, in an indirect water/glycol liquid-cooling
axial distance. system, channel height is not a sensitive factor affecting
System response parameters from different prediction heat transfer coefficient at the cell surface as it is in air
methods are presented in Table 2. Even though ' P is systems, even though water/glycol channel friction loss
predicted to be a little bit higher in CFD analysis due to is not as significant in magnitude as in air for a given
the entrance effect, the parameters are generally well coolant mass flow rate. But it is still recommended in a
matched between the prediction methods except for the liquid-cooling system to use a small gap coolant
maximum cell surface temperatures. The disagreement channel to reduce system weight and volume by
of ' Tmax mainly originates from the fact that CFD
analysis can capture the axially decreasing heat flux Table 2 System response parameters - Two
from cell to air, which makes the axial gradient of the different prediction methods
cell surface temperature smaller than that of the air 'P [Pa] 'T1 [ oC] 'Tmax [oC] 'T cell [oC] h [W/m2K]
temperature. This is because high conductivity channel
pressure
coolant
temperature
maximum
cell surface
maximum
cell internal
mean heat
transfer
materials inside a cell, such as the aluminum can, loss change temperature temperature coefficient
to inlet air to inlet air
would transfer internal heat flow to make cell Fully Developed Flow Relations
109.1 1.49 3.64 N/A 59.24
temperature more even. This result strongly implies that with Constant Heat Flux

capturing the internal heat flow paths and thermal CFD 114.2 1.50 2.89 3.41 60.96
resistances inside a cell are important for the improved

© 2007 WEVA Journal, pp. 126-133

130
 ISSN 2032-6653
The World Electric Vehicle Association Journal, Vol. 1, 2007

minimizing the amount of coolant in a system operated water/glycol

at a given coolant flow rate. CFD analyses were carried
out for a direct air-cooling system operated at (Dh , air
m c ) = (2.2 mm, 1.33 g/s) and for a water/glycol-jacket jacket
cooling system at (Dh , m  c ) = (4.0 mm, 1.70 g/s). Inlet aluminum can aluminum can
coolant temperatures were set at 35o C. The results cell core cell core
indicate that due to the large differences of coolant
velocity and kinematic viscosity between the two
systems, the air-cooling system needs to compensate for
a much larger pressure loss than the water/glycol
liquid-cooling system (221 Pa versus 7.68 Pa).
Fig. 12 Comparison of temperature distribution
Temperature distribution contours are shown for
contours between air cooled cell and
comparison in Fig. 12. Even though the rate of heat
water/glycol cooled cell
removal from cell to coolant is the same for the two
compared systems, airflow is rapidly heated compared Coolant Mean Temperature, AIR
Cell Surface Temperature, AIR

with water/glycol due to the small heat capacity of air Cell Temperature at center, AIR
Axis, AIR
Coolant Mean Temperature, WATER/GLYCOL
(a) Cell Surface Temperature, WATER/GLYCOL (b)
flow. In addition, the difference between coolant mean 41
Cell Temperature at center,
Axis, WATER/GLYCOL
WATER/GLYCOL
200
AIR
Channel Surface Temperature, WATER/GLYCOL 180
temperature and cell surface temperature is larger in the 40
160
WATER/CLYCOL

Heat Transfer Coefficient (W/m2K)

air-cooling system because of the smaller heat transfer 39
140

Temperature (oC)
coefficient. Therefore, not only the maximum 38
120

100
133.2 W/m2K

temperature but also the temperature nonuniformity 80 56.1 W/m2K

37
inside a cell are larger in the air-cooling system than in 60

40

the water/glycol system. Temperature contour-lines are 36

20

distributed similarly inside a cell in both cases. This 35

0 2 4 6 8 10 12
Axial Distance (cm)
14 16 18 20
0
0 2 4 6 8 10 12
Axial Distance (cm)
14 16 18 20

implies that internal heat flows are similar in both

systems, and consequently, that the internal heat paths Fig. 13 (a) Axial profiles of temperatures, and (b)
and the thermal resistances are determining the relative heat transfer coefficients at cell surfaces
temperature distribution inside a cell.
Axial temperature profiles of the air-cooling system uniformity. The CFD simulation results shown in Fig.
and the water/glycol system are compared in Fig. 13 (a). 13 are simply extended to a six-cell, serial-cooling
Compared with the water/glycol liquid cooled cell, the system and presented in Fig. 14. Other heat exchanges
maximum temperature on the cell axis in the air-cooled among the cells, such as conduction through electric
cell appears near the channel exit. A larger temperature connectors, are ignored. Cell temperature differences in
difference between each end of the cell shifts the a module more rapidly increase in the air-cooling
maximum temperature location further downstream. system proportional to the number of cells connected
Heat transfer coefficients at cell surfaces are plotted in serially with the coolant channel. Cell-to-cell
Fig. 13 (b) for both systems. temperature imbalance mainly comes from the coolant
temperature change in coolant channels. Due to the low
q cc (5)
h heat capacity of air, it is difficult to accomplish
Tcell _ surf  Tcoolant temperature uniformity inside a cell or between the
Thermal resistances of the jacket wall and air-gap cells in a module using air for cooling large or
layer, between the channel coolant and the cell surface, high-heat generating cells.
were considered to evaluate the heat transfer coefficient
of the water/glycol cooling system. Temperature 60
Cell Centerline Temperature, AIR
displacement between the channel surface and the cell Cell Centerline Temperature, WATER/GLYCOL
Coolant Temperature, AIR
surface in the water/glycol cooled cell shown in Fig. 13 55
Coolant Temperature, WATER/GLYCOL
(a) is due to these added thermal resistances. The higher
heat transfer coefficient of the water/glycol system
Temperature (oC)

50
leads to a smaller temperature difference between the
coolant and cell surface as shown in Fig. 13 (a) and Eq.
45
5. Even though the cell temperatures are distributed in
different ranges for each system, the magnitudes of
temperature differences between the cell center axis and 40

the cell surface of the air-cooled cell and the

water/glycol cooled cell are similar because the same 35
0 20 40 60 80 100 120
cell specifications are assumed for both cases. Cooling Channel Distance (cm)

Although, a serial-cooling system may provide

simplicity over a parallel cooling system because of for Fig. 14 Extended temperature profiles
its simpler of manifolding flow distribution, parallel for a six-cell serial cooling system
cooling provide a better cell to cell temperature

© 2007 WEVA Journal, pp. 126-133

131
 ISSN 2032-6653
The World Electric Vehicle Association Journal, Vol. 1, 2007

50
Transient Analysis In order to investigate the Air Cooling
45 Mineral Oil Direct Cooling
time-dependent thermal response of battery cooling Water/Glycol Jacket Cooling
40
systems, transient CFD analyses were carried out with a
cell specified in Fig. 8. An air-cooling system; a 35

Heat Transfer RAte (W)

mineral oil, direct liquid-cooling system; and a 30

water/glycol jacket cooling were compared. The 25 internal heat generation rate
systems were operated with the same channel geometry 20
and coolant mass flow rate, (Dh , m  c ) = (2.2 mm, 1.33 15
g/s). Initially, each system was in steady state with a
10
heat generation rate of 2 W. A system was heated with
5
sudden heat generation (50 W) for 2 min. Then, the cell
was cooled down to the initial steady-state conditions. 0
0 5 10 15 20 25 30 35 40
Coolant inlet temperatures were kept constant at 35 o C. Time (minute)

Using steady state fully developed relationships Fig. 15 Time variation of heat transfer
shown in Eq.1, channel pressure losses (' P) are rates from the cell to the coolant
predicted as 109.1 Pa, 418.3 Pa, and 18.27 Pa for the
50
air system, the mineral oil system, and the water/glycol 49 Cell Core Mean Temperature, AIR
system respectively. On the other hand, heat transfer 48
Coolant Outlet Mean Temperature, AIR
Cell Core Mean Temperature, MINERAL OIL
coefficients are predicted as 59.24 W/m2 K for the 47 Coolant Outlet Mean Temperature, MINERAL OIL

air-cooled surface, 318.2 W/m2 K for the mineral oil

46 Cell Core Mean Temperature, WATER/GLYCOL
45 Coolant Outlet Mean Temperature, WATER/GLYCOL
cooled surface, and 150.8 W/m2 K for an effective heat

Temperature (oC)
44

transfer coefficient at the water/glycol jacket cooling 43

42
cell surface. The mineral oil, direct-contact, 41
liquid-cooling system is expected to have a much 40

higher heat transfer coefficient than the other coolant 39

38
systems (direct air and indirect water/glycol) when it is 37
operated at a small channel (Dh ) and large coolant flow 36

 c ) at the expense of high pressure loss in the

rate ( m 35
0 5 10 15 20 25 30 35 40
Time (minute)
coolant channel. The water/glycol system has a larger
heat capacity flow rate ( m  c c p , 4.42 J/Ks) than the Fig. 16 Time variations of mean
mineral oil system (2.53 J/Ks) or airflow system (1.34 temperature of the cell core and the
J/Ks), which means it can remove the suddenly released coolant outlet temperature
heat from the cell with a smaller increase of coolant
temperature. The simulation results showing the water/glycol cooled cell, even though the coolant
transient effect of using high h and a high heat capacity temperature of mineral oil system is higher than the
coolant system are presented in Fig. 15 and Fig. 16. temperature of water/glycol. The results imply that a
The heat transfer rates from the cell to the coolant are high h system (and high m  c c p system, if possible) is
compared in Fig. 15. Due to the highest heat transfer preferred for limiting the maximum peak temperature of
coefficient, released heat is most quickly removed from a cell and damping out the temperature oscillation in
the mineral-oil-cooled cell. On the other hand, the heat highly transient heat generating battery systems.
rejecting amount in water/glycol system is comparable
with the mineral oil cooling system in spite of its 3. CONCLUSIONS
smaller heat transfer coefficient. This is because the
temperature potential is larger in the water/glycol To achieve performance and cost-effective cooling of
system for a lower coolant temperature. In a high h and a battery module/pack for electric or hybrid vehicles,
high m  c c p system, a large amount of heat can be the system thermal responses and their sensitivities are
transferred from cell to coolant with a small evaluated as a function of controllable system
temperature increase of the cell while the coolant parameters (e.g., Dh and m  c ). For given cell
temperature does not change much. specifications, different types of coolant are examined:
air cooling, direct-contact liquid cooling using mineral
Time variations of the mean temperature within the oil, and (indirect) water/glycol jacket cooling.
cell core and the mean temperature of the coolant outlet The simplicity of an air-cooling system is an
are shown in Fig. 16 for the air-, mineral-oil-, and advantage over a liquid-coolant system. Air cooling
water/glycol-jacket cooling systems. Since the heat could have less mass, has no potential for leaks, needs
rejection rates of both liquid-cooling systems are fewer components, and could cost less. However, the
similar to the given conditions, as it is shown in Fig. 15, heat transfer coefficient (h) of an air-cooling system is
the average cell temperature variation profiles are also lower than that of other coolant systems. Another
similar. In both the liquid-cooling systems, the peak drawback of an air system comes from its small heat
temperature is lower and the cooldown time is shorter capacity. Due to the small heat capacity of air, it is
than in the air-cooling system. Note that the mineral oil difficult to accomplish temperature uniformity inside a
system cooled down a little bit faster than the cell or between the cells in a module for large format

© 2007 WEVA Journal, pp. 126-133

132
 ISSN 2032-6653
The World Electric Vehicle Association Journal, Vol. 1, 2007

application. The temperature difference between and Tien Duong of DOE

coolant air and cell surface (' T2 ) is sensitive to
variations of Dh due to small heat conductivity of air.
The heat transfer coefficient (h) is inversely REFERENCES
proportional to Dh, while pressure loss in channel (' P) [1] W.B. Gu and C.Y. Wang, "Thermal and Electrochemical
is inversely proportional to Dh 3. Therefore, increasing h Coupled Modeling of a Lithium-Ion Cell, in Lithium
Batteries ,´ ECS Proceedings, Vol.99-25 (1), pp.748-762,
by reducing Dh is limited by the required blower power. 2000.
Reducing coolant air temperature change inside a [2] V. Srinivasan and C.Y. Wang, ³Analysis of
system (' T1 ) by increasing flow rate ( m  c ) is also Electrochemical and Thermal Behavior of Li-Ion Cells,´
limited by blower power and size. Journal of Electrochemical Society, Vol. 150,
ppA98-A106, 2003.
Liquid-cooling is more effective in heat transfer and [3] A. Pesaran, D. Bharathan, G.-H. Kim, A. Vlahinos, and
takes up less volume, but the added complexity and cost T. Duong, ³Improving Battery Design with
may outweigh the merits. Maintenance and repair of a Electro-Thermal Modeling´, in 21st Electric Vehicle
Symposium, Monte Carlo, Monaco, 2005.
liquid cooled pack is more involved and costlier. [4] D. Bharathan, A. Pesaran, G.-H. Kim, and A. Vlahinos,
Indirect liquid cooling, with jackets, is easier to handle ³Electro-Thermal Modeling to Improve Battery Design´,
than direct liquid cooling. IEEE Vehicle Power and Propulsion Conference, IEEE,
A 50/50 water and glycol mixture for an Chicago, IL USA, 2005.
[5] S. Al-Hallaj, et. al, ³Novel PCM Thermal Management
indirect-cooling system, selected because of its low Makes Li-Ion Batteries A viable Option for High Power
freezing temperature for vehicle applications, has much Applications,´ Battery Power Products and Technology
lower viscosity than dielectric mineral oil for direct Magazine, November 2004.
liquid cooling. Therefore, increasing the coolant flow
rate may not be as severely restricted by the pump BIOGRAPHIES
power as it is in a mineral oil direct-cooling system.
Water/glycol has a higher heat capacity. So, the coolant
Gi-Heon Kim has worked with the Advanced
temperature change inside a system can be greatly
Vehicle Systems group at NREL since 2004. His
reduced by using water/glycol as a heat transfer fluid in
recent research interests in advanced vehicle
the system. This means that cell/module temperature
energy storage system tasks include development
uniformity can be effectively achieved even in a serial
of a three-dimensional Li-Ion battery thermal
cooling system if the coolant paths are properly
abuse model and hybrid electric vehicle//electric
designed. Water/glycol mixtures generally have a
vehicle battery thermal management system
higher thermal conductivity than oil. However, due to
modeling. He has been working on various
the added thermal resistance between coolant and cell
vehicle power system research topics, such as PEM fuel cell freeze
surface, such as jacket wall and air gap, the effective
start and alternative fuel based diesel after-treatment systems. His
heat transfer coefficient at the cell surface is greatly
expertise is in heat and mass transfer, fluid mechanics, turbulence,
reduced. Because of the added thermal resistances, h is
not as sensitive to the variation of Dh . combustion, reacting flows and computational fluid dynamics. Dr.
A mineral oil direct-contact liquid-cooling system Kim completed a Ph.D. in Mechanical Engineering from Colorado
has a much higher heat transfer coefficient (h) than the State University with research on performance improvement of
other coolant systems when it is operated at a small natural gas engines. In addition to his Ph.D., Dr. Kim holds M.S. and
channel (Dh ) and a large coolant flow rate ( m  c ) at the B.S. in Mechanical Engineering from Seoul National University in
expense of high pressure loss in coolant channel. So, a Seoul, Korea.
mineral-oil-cooling system may be preferred for Ahmad Pesaran joined NREL in 1983 and has
limiting the maximum peak temperature of a cell and been working on various energy systems, such as
damping out the temperature oscillation in certain solar cooling, ocean thermal energy conversion,
circumstances, such as in highly transient heat air conditioning, desiccant dehumidification/
generating battery systems. cooling for buildings and buses, and most recently
CFD analysis captured the axially decreasing heat hybrid electric vehicles. Since 1995, he has been
flux from cell to air, which makes the axial gradient of working on hybrid electric vehicle projects. He is
cell surface temperature smaller than that of air currently the project manager for various activities related to battery
temperature. This implies that capturing the internal thermal characterization, battery thermal analysis, and battery
heat flow paths and thermal resistances inside a cell modeling and management. Dr. Pesaran holds a B.S. in chemical
using a sophisticated three-dimensional cell model is engineering from Shiraz University, as well as an M.S. in engineering
important for the improved prediction of cell/battery and a Ph.D. in mechanical engineering from UCLA. He is a member
thermal behaviors. of the FreedomCAR Electrochemical Energy Storage Technical Team.

ACKNOWLEDGEMENT
The U.S. Department of Energy (DOE) FreedomCAR
and Vehicle Technologies Program funded this effort.
We appreciate the support provided by Dave Howell

© 2007 WEVA Journal, pp. 126-133

133