1 s2.0 S0017931023008931 Main

International Journal of Heat and Mass Transfer 218 (2024) 124748
Contents lists available at ScienceDirect
International Journal of Heat and Mass Transfer

journal homepage: www.elsevier.com/locate/ijhmt
Review of thermal coupled battery models and parameter identification for

lithium-ion battery heat generation in EV battery thermal
management system
Jie Liu a, 1, Saurabh Yadav b, 1, Mohammad Salman a, Santosh Chavan c, *, Sung Chul Kim a, *
a
School of Mechanical Engineering, Yeungnam University, 280 Daehak-ro, Gyeongsan-si, Gyeongbuk 712-749, South Korea
b
School of Mechanical Engineering, VIT University, Chennai Campus, Chennai, Tamil Nadu 600127, India
c
Department of Information and communication, Yeungnam University, 280 Daehak-ro, Gyeongsan-si, Gyeongbuk 712-749, South Korea
A R T I C L E I N F O A B S T R A C T
Keywords: With the increasing popularity of electric vehicles, there is a growing need for efficient and reliable thermal
Battery modeling management systems to regulate the temperatures of lithium-ion batteries. Heat generation, which can lead to
Electric vehicle decreased performance, reduced lifespan, and even safety hazards if not properly managed, is a major challenge
Battery thermal management system
for these batteries. This review provides an overview of the numerical modeling techniques used to simulate heat
Equivalent circuit model
Electrochemical model
generation and thermal management in lithium-ion batteries for electric vehicles. Various approaches to
Parameter identification modeling the thermal behavior of batteries are reviewed with regard to thermal coupled battery models for the
thermal analysis of battery and battery thermal management system (BTMS). This review discusses various
modeling approaches, including simple battery models, electrical circuit models (ECMs), and physics-based
electrochemical models, focusing on the Newman P2D model. Thermal runaway model that is employed to
inform safety design for lithium-ion battery is also reviewed. In addition, the parameter identification methods
for these models are reviewed. The merits and drawbacks of each model are discussed in detail, and the com
plexities involved in parameter identification are investigated. This comprehensive review aims to guide re
searchers and engineers in selecting and utilizing appropriate tools for battery thermal management studies.
1. Introduction maximum amount of electricity that a battery can store and deliver; this
capacity reduces with the use of electrodes or electrolytes. C-rate refers
Currently, almost 80 % of the global energy supply depends on fossil to the charging or discharging rate and is used to measure the discharge
fuels, such as coal, oil, and natural gas. Most large-scale production and of a battery relative to its maximum capacity [4]. The power perfor
consumption of energy are believed to result in environmental pollution mance and driving range of electrical vehicles (EVs) and hybrid EVs are
and adverse effects on human health [1,2]. Owing to the world’s significantly influenced by the type of battery used as their primary
increasing reliance on renewable energy sources, electric vehicles (EVs) power source. Several battery types, including lead-acid, nickel-based,
present the most promising solution for achieving a cleaner and more sodium-based, and lithium-ion batteries (LIBs), have been developed
sustainable environment because batteries provide a superior means of [5]. Among these, LIBs are the most preferred energy storage device for
storing consistent energy generated from these sources [3]. electric vehicles (EVs) because of their superior characteristics,
Batteries are devices that utilize an electrochemical reaction between including higher energy density, specific power, recyclability, lighter
two or more cells to provide power to electrical devices and possess weight, longer cycle life, lower self-discharge rates, and shorter charging
several common features, such as auto-discharge rate, specific energy, time [5,6].
operational conditions, toxicity, charging time, and C-rate. Battery ca Despite technological advancements, thermal safety concerns related
pacity, which is determined by the reversibility conditions of the elec to Li-ion batteries in EVs continue to pose significant challenges, making
trochemical process, is a crucial characteristic because it represents the them a crucial area of research and development. An effective battery
* Corresponding authors.
E-mail addresses: [email protected] (S. Chavan), [email protected] (S.C. Kim).
1
These authors contributed equally to this work.
https://doi.org/10.1016/j.ijheatmasstransfer.2023.124748
Received 21 May 2023; Received in revised form 2 July 2023; Accepted 22 September 2023
Available online 3 October 2023
0017-9310/© 2023 Elsevier Ltd. All rights reserved.
J. Liu et al. International Journal of Heat and Mass Transfer 218 (2024) 124748
thermal management system (BTMS) for EV batteries is urgently needed phase-change material (PCM)-based cooling [23–25,10], and hybrid
to ensure the safety of the vehicle, battery, and its occupants and cooling [26,27]. Although some numerical studies are included in the
improve battery lifetime and operation under different environmental abovementioned literature, a discussion of the numerical models and
conditions [7]. The correlation between the operating temperatures and methodologies used in these numerical investigations is not presented in
the performance and safety of lithium-ion batteries has been widely depth. However, numerical modeling of BTMS is a significant approach
recognized. Fig. 1 demonstrates that the ideal operating temperature for in the development of battery technology [28]. It is necessary to conduct
these batteries falls between 15 and 35 ◦ C. If the temperature during the a comprehensive review of numerical methods for BTMS. Therefore, this
charging or discharging process exceeds this acceptable range, the reg study reviews various approaches for modeling the thermal behavior of
ular functioning of the batteries can be compromised. Additionally, the batteries with regard to thermal coupled battery models and parameter
temperature difference between connected LIB cells in the battery identification for BTMS.
module or pack should be less than 5 ◦ C [8–11]. The purpose of a BTMS
is to control temperature levels and distribution, which is essential for 2. Overview of battery models and battery heat generation
extending the lifespan of lithium-ion cells and the overall battery sys
tem. To achieve this goal, the system utilizes various techniques, such as Numerous battery models with varying complexities that offer utility
monitoring battery health to determine the state of charge (SOC) and across a range of applications have been developed. According to the
state of health (SOH) and providing relevant battery information to EV level of physical insight, battery models are divided into three types:
users. Furthermore, the BTMS ensures safe and efficient battery opera white box, gray box, and black box models, which are represented by the
tion by maintaining an appropriate operating temperature within the electrochemical model, equivalent circuit model (ECM), and neural
required range [11]. However, studies on BTMS have focused on battery network model, respectively [29]. Fotouhi et al. summarized battery
cooling technologies to ensure battery performance and safety, and models into mathematical models, ECMs, and electrochemical models
relatively little attention has been paid to preheating [12]. Research on and investigated their uses within EV battery management systems [28].
BTMS can be categorized into two general approaches: direct mea In this study, the review approach for EV Li-ion battery models focuses
surement and indirect calculation. Direct measurement methods involve on commonly used modeling methods for the thermal performance of
experimental analysis of heat production characteristics; however, they Li-ion batteries. Therefore, the neural network model mostly used for
are time-consuming, costly, and may not fully capture the comprehen battery life prediction and the mathematical model that is no longer
sive thermal state and temperature distribution of batteries. By contrast, used are not discussed. Generally, during the charging or discharging
indirect calculation methods rely on battery numerical models to process, heat generation can be estimated by the sum of irreversible and
analyze the internal state of a battery, which may be challenging to reversible heat (entropic heat) and is given as [30]
measure experimentally. Despite these limitations, indirect calculation
qgen = qirr + qrev (1)
methods offer a more efficient and cost-effective approach to evaluating
the thermal characteristics of lithium-ion batteries [13]. The use of Both irreversible and reversible heat are generated by various
numerical methods is crucial for predicting the safety and performance chemical and physical processes inside the battery. Irreversible heat is
of batteries in different end-use contexts. This is particularly important often caused by the flow of current through the battery’s internal
in designing new batteries, where measurement data obtained from resistance, resulting in Joule heating, polarization effect, etc.
previous experiments can be used to validate predictive models and Conversely, reversible heat is generated or absorbed during the charging
achieve accurate simulation data [14]. and discharging processes owing to the entropy changes of the active
Numerous scholars have conducted review studies on BTMS for EVs, materials. Therefore, based on the methodology for representing the
focusing on various perspectives. Most of the reviews are comprehensive chemical and electrical features used to evaluate battery heat genera
discussions about the status and development potential of BTMS and a tion, the following three types of numerical models are discussed in this
comparison of the advantages and disadvantages of various BTMS study:
technologies at the technical level [1,3,6–8,12,15–17]. BTMS, particu
larly for cooling technology, can be classified into three categories based 1. Simple battery model
on the use of parasitic power: active, passive, and hybrid cooling. Active 2. Equivalent circuit model (ECM) model
cooling comprises cooling systems that use air or liquid and some 3. Physics-based electrochemical model
emerging approaches. Passive cooling, on the other hand, utilizes
techniques such as phase change materials (PCM), heat pipes, and nat Parameter identification is a key aspect for battery modeling.
ural convection to dissipate heat. Finally, hybrid cooling uses a combi Therefore, after introducing the battery models, parameter identifica
nation of two or more cooling strategies and can be categorized as tion using experimental methods is reviewed for the different models.
air-PCM, air-heat pipe, heat pipe-PCM, and liquid-PCM systems.
Accordingly, review studies on different types of lithium battery cooling 3. Simple battery model
technologies have also been presented, such as air cooling [9,11,18,9],
liquid cooling [19], immersion cooling [20], heat pipe [21,22], Simple battery models require fewer computational resources and
Fig. 1. Relation between battery performance and operating temperature (reproduced from Ref. [8]).
2
less processing time than detailed electrochemical or multiphysics total heat generation in LIBs. The entropy change of Li+ ions during
models. This makes them suitable for battery management systems discharge or charge processes depends on the temperature variation of
(BMS), where rapid calculations are necessary to ensure optimal oper the open-circuit voltage E0 and can be expressed as follows [33]:
ation and safety. As simple battery models have fewer parameters and
dUOC
equations, they are easier to implement and understand. This can be ∇S = nF (6)
dT
beneficial to engineers and researchers who may not have extensive
expertise in electrochemistry. A simple battery model (or theoretical where n and F are the number of electrons exchanged per mole and the
resistance model), as shown in Fig. 2, consists of an internal resistance Faraday constant, respectively. Therefore, the Eq. (3) and (4) can be
(R0), an open-circuit voltage (UOC), and a final battery voltage (Ub) and expressed as [33]
assumes that the heat generation rates inside the battery have a uniform [ ]
distribution. It is also known as the internal resistance model or Rint qgen =
Ib
(UOC − Ub ) − T
∇S
(7)
model because the battery’s internal resistance is always denoted as Vb nF
“Rint.” [ ]
The dynamics of a battery cannot be described using the simplified 1 2 ∇S
qgen = Ib R0 − Ib T (8)
model because the polarization effect is not included. Consequently, the Vb nF
internal resistance equivalent model is commonly used in basic battery The simple battery thermal models used in these studies do not
simulation analyses. However, their practical applications are infre reflect the internal electrochemical characteristics of the battery, such as
quent because of their inherent limitations. The terminal voltage of the the polarization effect. Instead, they rely on experiments to evaluate Ub,
battery can be expressed as R0, d UOC /dT, or ▽S/nF as functions of T or SOC, or both of them, and
Ub = UOC − Ib R0 (2) the internal current I of the battery is set to be a constant value based on
the charging or discharging rate (C-rate). Here, the battery SOC as a
In the simplified model, the heat transfer phenomena in the battery function of the internal current I, battery normal capacity, and time can
can be simplified to thermal diffusion with a heat source, and the energy be calculated as [34]
balance equation inside the battery is as follows:
I(t)
SOC(t) = SOC(t − 1) + Δt (9)
∂T ∂2 T ∂2 T ∂2 T Capacity
ρcp = λx 2 + λy 2 + λz 2 + qgen (3)
∂t ∂x ∂y ∂z
Owing to the low computational resources required, the simple
where ρ, cp, T, t, λ, and q are the battery density, specific heat, tem battery model is commonly coupled with three-dimensional (3D)
perature, time, thermal conductivities, and heat generation rate per unit computational thermal models to address complex geometric systems or
volume of the battery, respectively. Eq. (1) can be expressed as an en BTMS issues that require extensive calculations [34–61]. Models of
ergy balance equation as follows [31,32]: thermal management systems either contain more complex geometries,
[ ] such as cooling channels or cooling plates for liquid cooling, or more
qgen =
Ib
(UOC − Ub ) − T
dUOC
(4) complex computational models, such as phase change phenomena in
Vb dT PCM or heat pipe cooling. Commercial computational fluid dynamic
(CFD) packages, such as ANSYS Fluent and COMSOL, are the best tools
where I, UOC, Ub, T, Vb, and d UOC /dT are the current passing through the for solving this coupling model. In these models, the electrical field and
battery, battery open-circuit voltage, battery voltage, battery tempera chemical reactions are not included, and the battery cell or packed
ture, battery volume, and entropy thermal coefficient, respectively. The module is defined as a solid body with a heat source, which is calculated
1st and 2nd terms on the right side of Eq. (3) represent the generated using Eq. (7) or (8). Jiahao Cao et al. [43] developed a thermal coupled
irreversible and reversible heat, respectively. Because the internal simple battery model for a liquid cooling with PCM hybrid BTMS by
resistance of the battery can be expressed as R0=(UOC-Ub)/I, Eq. (4) can using ANSYS Fluent. The model was validated by the experimental data.
also be expressed as follows [32]: From Fig. 3, it can be seen that the simulation results can visualize the
1 2
[
dUOC
] temperature distribution of the battery and the BTMS system. However,
qgen = Ib R0 − Ib T (5) the electrical and chemical characteristic, such as current density,
Vb dT
electrode potential, of the battery cannot be represented in this nu
Eq. (4) was developed by Bernardi and has been widely used as an merical study.
energy-governing equation in simplified battery thermal models. Table 1 lists some typical studies that utilized commercial CFD
Reversible heat, also called entropy heat, contributes significantly to the solvers to solve battery thermal problems using simple battery models
coupled with 3D thermal models. Based on the conclusions of these
studies, most of them pay more attention to the design of thermal
management systems and the improvement of cooling performance,
disregarding the electrochemical performance. Most battery thermal
management systems, including air cooling, liquid cooling, PCM, HP,
and hybrid cooling, can be solved using this coupled model. These
models generally have the following common assumptions [32,34,35]:
(1) The material properties of the battery are assumed to be inde

pendent of temperature and SOC owing to their significantly
small variations.
(2) The heat generated inside the battery is uniformly distributed.
(3) The charging or discharging rate is constant.
(4) The radiation heat transfer inside the battery is ignored.
Based on these assumptions, the transport phenomena inside the

battery can be simplified to thermal diffusion with heat generation, as
Fig. 2. Simple battery model.
3
Fig. 3. Temperature distributions of battery with different 25wt % and 67wt % PCM [43].
described by Eq. (3). However, more complex models, such as the ∫t ( )

equivalent circuit model (ECM), single particle model (SPM), or pseudo- Ui (t) =
Ib
−
Ui
(24)
two-dimensional (P2D) model, which can capture a wider range of C i Ri C i
t0
battery behaviors, are needed for more accurate modeling of battery
behavior.
n ∫ ( )
t
∑ Ib Ui
Ub = UOC − Ib R0 − − (25)
4. Equivalent circuit model (ECM) coupled thermal model i=1
C i Ri C i
t0
Although electrochemical battery models based on physical science Because the polarization effect is considered in ECM, the battery heat
can accurately describe the dynamics of the particles within a battery, generation can be represented as [88]
one of their drawbacks is their complexity, necessitating the identifi qgen = qp + qj + qr + qs (26)
cation of numerous parameters, which is considered a challenging
process. Utilized extensively in battery modeling, simulation, and SOC where qgen, qp, qj, qr, and qs are the polarization, Joule, reaction, and side-
estimation, ECMs achieve an optimal balance between precision and reaction heats, respectively. qs is the heat generated during the self-
complexity. The physical variables within ECMs are not only readily discharge process and can be ignored owing to its significantly small
interpretable but also comprehensible. value during the charging or discharging of the battery [88]. Therefore,
ECM is also a simplified representation of a battery using electrical Eq. (4) can be represented as
components such as resistors, capacitors, and voltage sources. Fig. 4 ⎡ ⎤
presents four commonly used ECM models: 1st-order Thevenin model, n ∫ ( )
t
Ib ⎣ ∑ Ib Ui dUOC ⎦
2nd-order Thevenin model, the partnership for a new generation of qgen = Ib R0 + − − T (27)
Vb Ci Ri Ci dT
vehicle (PNGV) model, and the general nonlinear (GNL) model [62].
i=1
t0
Unlike the simple model, ECM does not only consist of a simple Yubai Li et al. [82] focused on the three-dimensional (3D) thermal
voltage source and internal resistance. As shown in Fig. 4, the Thevenin modeling of a single Lithum-ion battery cell and a 50 V Lithum-ion
models are composed of one (1st order) or two (2nd order) RC network battery pack comprising 14 prismatic batteries. The electrochemical
structures to consider the polarization effects of the internal chemical modeling was conducted at a subscale using an equivalent circuit model.
reaction of the battery [63]. The 1st-order Thevenin model is obtained From the simulation results, it is observed that, unlike simplified models,
by adding one RC network to the simple internal resistance model to the thermal model based on the equivalent circuit model (ECM) not only
represent the polarization effect and better describe the charging or accurately captures the temperature distribution (Fig. 5a) of the battery
discharging process. The 2nd-order Thevenin model, also called the but also captures the electrochemical characteristics such as electrode
dual-polarization (DP) model, is obtained by adding two parallel RC potential (Fig. 5b) and battery voltage (Fig. 5c).
networks to the Rint model. Lastly, the PNGV and GNL models can be More related studies on thermal coupled modeling using ECM are
obtained by adding a series equivalent capacitance C0 to the 1st-order summarized in Table 3. In these studies, the Thevenin and DP models are
Thevenin and DP models, respectively. The equivalent capacitance C0 mostly used, whereas the PNGV and GNL models have never been
in the PNGV and GNL models is used to describe the changes in the adopted because the changes in the open-circuit voltage caused by time
open-circuit voltage (OCV) caused by time accumulation. The governing accumulation, which is described by the equivalent capacitance C0, are
equations for the battery behaviors of the four models are summarized in almost not considered in the battery thermal analysis. The Thevenin
Table 2. model is predominantly used to evaluate the thermal characteristics of
Generally, the PNGV and GNL models are more accurate than the the batteries. Therefore, it can be observed that most of the studies [64,
Thevenin models in representing the electrical and chemical behaviors 66–68,73] using this model are usually only coupled with the lumped
of the battery. However, the latter is mostly used for modeling and thermal model and do not involve complex battery thermal management
evaluating battery heat generation. Some studies have presented bat systems. Regarding the strategies for the design, development, and
teries with more than two RC networks [84–87]. Therefore, the gov enhancement of BTMS, the most commonly used numerical method
erning equations for the Thevenin models can be summarized as follows involves coupling a DP battery model with a CFD model [74,77,79,80,
[74,84]: 82,85]. As previously discussed, in the coupling of the Rint coupled CFD
∑
n model, the Rint model merely serves as an auxiliary tool for determining
Ub = UOC − Ib R0 − Ui (22) battery heat generation. In contrast, ECM, with its complexity, provides
a more comprehensive representation of battery behavior, incorporating
i=1
dUi Ui Ib multiple resistive and capacitive elements to better capture dynamic

=− + (23) behavior. The Rint model simplifies this with a single resistance
dt Ri C i C i
4
Table 1 Table 1 (continued )

Studies on thermal coupled modeling using simple battery model. Author Battery Cooling Software Conclusion
Author Battery Cooling Software Conclusion type method
type method
the EV dynamics
Wang et al. [32] Module (10 Air/PCM Fluent Five thermal is developed to
pouch management investigate the
cells) methods are thermal
numerically behaviors of Li-
investigated and ion battery
compared. A fin operating under
structure can dynamic battery
enhance the loads.
cooling effect, E et al. [42] Module (20 Air Fluent The cooling
and the PCM can cylindrical performance of
greatly improve cells) inlet and outlet
the temperature located on the
uniformity of different sides is
the battery cell. better than on
Akbarzadeh Module (12 Air/Liquid Fluent The liquid type the same side
et al. [34,37] prismatic BTMS results in and the use of
cells) a lower module the baffle plate
temperature and can highly
better improve the
temperature thermal
uniformity. performance of
Chen et al. [35] 1 pouch Liquid Fluent The optimized air cooling
cell design cooling strategies with
plate can much the lateral inlet
reduce the cell and outlet.
temperature and Cao et al. [43] Module (20 Hybrid Fluent A numerical
control the cylindrical (Liquid+PCM) model has been
temperature cells) created to
deviation within analyze the
a small range. hybrid thermal
Behi et al. [36] Module (24 Hybrid COMSOL The cooling management
cylindrical (Air+HP) strategy using system, and has
cells) heat pipe with a been verified
copper presents through a
the best cooling comparison
performance for with
both of experimental
temperature data. The model
reduction and will serve as a
temperature guide for
uniformity. enhancing the
Liu et al. [38] 2 prismatic Hybrid Fluent The designed structure,
cells (Liquid+PCM) hybrid cooling control mode,
system presents and overall
more than 30 K performance of
temperature the hybrid
reduction than BTMS.
the nature Kausthubharam Module (9 Air COMSOL Replacing the
convection. et al. [46] cylindrical air inlet-outlet
Wang et al. [39] Module (20 Liquid Fluent Increasing the cells) boundaries to a
cylindrical coolant mass central location
cells) flow rate can improved the
greatly reduce temperature
the battery uniformity in
temperature and the battery
improve the module.
temperature Tan et al. [47] Module (64 Liquid Fluent The direct liquid
uniformity of cylindrical cooling with a
the module. cells) full-size channel
Lee et al. [40] Module (8 Hybrid Fluent The proposed design and high
pouch (Liquid+PCM) BTMS is not only flow velocity
cells) effective in control has a
terms of its redundant
cooling cooling
performance but performance
also sufficiently with the
usable in the problems of low
continuous energy density
cycle. and high power
Han et al. [41] Module (10 – Fluent The dynamic consumption,
prismatic battery thermal which can be
cells) model resolved
integrated with through
optimization
(continued on next page)
5
Table 1 (continued ) electrochemical processes and transport phenomena that occur in bat
Author Battery Cooling Software Conclusion teries during charge and discharge cycles. These models are particularly
type method useful for studying battery performance. Electrochemical-thermal
Liang et al. [49] Module (5 Liquid COMSOL Three important
coupled battery models offer an efficient and robust approach for
cylindrical system exploring the properties of lithium-ion batteries. These models typically
cells) parameters focus on a single battery unit that includes current collectors, positive
required electrodes, separators, and negative electrodes. By providing precise and
determining
reliable numerical outcomes, they eliminate the need for complex
moderately for
balancing the experimental analysis [89]. As the basis of other battery models, they
system cooling can better reflect the inconsistency inside a single battery cell [90].
performance, There are several physics-based electrochemical models for battery
power simulations, each providing different levels of detail and complexity,
consumption,
and lightweight.
such as the single particle model (SPM), Newman
Jang et al. [52] Module (10 Hybrid CFX The maximum pseudo-two-dimensional (P2D) model, and Doyle-Fuller-Newman (DFN)
prismatic (Liquid+HP) temperatures of model. SPM simplifies the battery representation by assuming that each
cells) the battery electrode can be described by a single representative particle. The model
module in the
focuses on the transport of lithium ions within the solid electrode par
liquid cooling
with B-type heat ticles and the reaction kinetics at the electrode/electrolyte interface.
pipes decrease This omits the transport phenomena in the electrolyte and assumes a
by 6.1 ◦ C and uniform concentration of lithium ions in the electrolyte [91]. Because
9.4 ◦ C under the the potential drop caused by the electrolyte dynamics is significant and
basic and
optimized
cannot be ignored, SPM is rarely used to evaluate battery heat genera
conditions tion, although some studies have improved the model by using ECM [91,
relative to those
in the only
liquid cooling,
respectively.
Table 2
Kang et al. [53] 1 pouch – COMSOL The acquired Governing equations for ECMs.
cell regional Models Equations References
resistance
integrates 1st order Thevenin Ub = Uoc − Ib R0 − U1 (10)
[64–73]
electrical
dU1 U1 Ib (11)
resistance and = − +
dt R1 C1 C1
effect of entropy
2nd order Thevenin Ub = Uoc − Ib R0 − U1e − U1c (12)
changes [74–82]
dU1e U1e Ib (13)
corresponding = − +
dt R1e C1e C1e
to the SOC. dU1c U1c Ib (14)
= − +
dt R1c C1c C1c
PNGV Ub = Uoc − Ib R0 − U1 − UC0 (15)
element. This complexity typically lends the ECM coupled with CFD a [83]
dU1 U1 Ib (16)
higher accuracy in predicting battery behavior under various operating
= − +
dt R1 C1 C1
conditions. Even in terms of thermal predictions, a detailed represen dUC0 Ib (17)
=
dt C0
tation of the internal battery processes by ECM might yield more accu GNL model Ub = Uoc − Ib R0 − U1e − U1c − UC0 (18) /
rate results. dU1e U1e Ib (19)
= − +
dt R1e C1e C1e
dU1c U1c Ib (20)
5. Physics-based electrochemical-thermal model dt
= − +
R1c C1c C1c
dUC0 Ib (21)
Physics-based electrochemical models for battery simulations are
=
dt C0
designed to provide a detailed understanding of the underlying
Fig. 4. Schematics of battery equivalent circuit models (ECMs): (a) 1st-order Thevenin model; (b) 2nd-order Thevenin model (DP model); (c) PNGV model; (d) GNL
model [62].
6
Fig. 5. Simulation results by the thermal model based on the equivalent circuit model (ECM). (a): battery temperature distribution under 5C discharging at 660 s
with different coolant velocities; (b): distributions of φ+ and φ− under 0.5C discharging. (c): ECM model parameters adjustment and comparison with experimental
data under various discharging rates [82].
Table 3
Studies on thermal coupled modeling using ECM.
Author Battery type ECM Thermal Cooling methods Conclusion
model model
Hariharan. Cylindrical cell Thevenin 1D Lumped – The suggested model, with a universal set of parameters, is proficient in representing
[64] cell voltage and temperature across a broad spectrum of power levels.
Calearo et al. 40 kW Pack Thevenin 1D Lumped Active cooling The model forecasts the electro-thermal behavior of the battery pack with a margin of
[66] error that stays under 5 %.
Wang et al. Pouch cell Thevenin 1D Lumped – The suggested hybrid modeling approach presents a universal framework that can
[67] accommodate additional influences like the health state and current within battery
models.
Estevez et al. Pouch cell Thevenin 3D Lumped – The proposed model can serve as an instrumental resource for establishing new
[68] module architectures and aiding in the design of battery cooling systems.
Liu et al. [72] Module (25 Thevenin 1D Lumped Liquid cooling The study discusses the impact of various factors on the thermal performance of the
cylindrical cells) battery module. These include two types of tube layouts, fluid velocity, tube
diameter, gradient ratio of flow velocity, and gradient increase in tube diameter.
Xu et al. [73] Module (20 Thevenin 1D Lumped – Simulation and experimental findings demonstrate that the prototype simulation
cylindrical cells) system is capable of not just delineating the electrical and thermal characteristics of
the battery, but also estimating the state of the battery with high precision.
He et al. [74] Module (12 DP 1D Hybrid cooling The introduced model is computationally effective and can be readily applied to
prismatic cells) Lumped/ (Liquid-Air-PCM) enhance algorithms for optimal temperature control during the operation of battery
CFD modules.
Hou et al. Module (3 DP 1D Lumped Hybrid cooling The model has potential for real-time forecasting of electrical and thermal signals in
[75] prismatic cells) (Liquid-PCM) large battery packs, along with its utility in analyzing and optimizing thermal
management systems.
Wang et al. Cylindrical cell DP 1D Lumped – The introduced model is capable of providing online estimations of both internal and
[76] external temperatures, and it does so with a high degree of accuracy.
Kirad et al. Module (30 DP CFD Air cooling The study reveals that the transverse spacing between the batteries significantly
[77] cylindrical cells) influences the cooling efficiency factor, while the longitudinal spacing impacts the
temperature uniformity.
Saqli et al. Module (3 DP CFD – Achieving real-time updates on the battery’s operating temperatures, capacity, and
[79] cylindrical cells) state of charge (SOC) has proven successful in enhancing the accuracy of estimations.
Behi et al. Module (30 DP CFD Evaporative The results indicate that the implementation of the evaporative cooling method
[80] prismatic cells) cooling/HP results in a significant temperature reduction in both the cell and module, with
decreases of 35.8 % and 23.8 % respectively.
Li et al. [82] Module (14 DP CFD Liquid cooling The developed model offers in-depth, 3D and time-resolved understanding of thermal
prismatic cells) dynamics and battery behavior under rapid discharge and harsh conditions, and is
broadly applicable to the design of individual Li-ion batteries as well as entire Li-ion
battery packs.
Lin et al. [84] Cylindrical cell Multi-RC 1D Lumped – The suggested scheme for parameterization enables individual identification of
electrical and thermal sub-models, substantially simplifying the complexity involved
in the process of parameterization.
Saw et al. Pack (24 × 28 Multi-RC 1D – The simulation results of the battery model align well with the experimental data, and
[85] pouch cells) Lumped/ show that an active thermal management system is necessary to enhance the
CFD longevity and ensure the safety of the battery pack
92]. The DFN model is an extension of the P2D model and accounts for The Newman P2D model is a physics-based model that uses porous
the detailed transport of lithium ions in the electrolyte using the electrodes and concentrated solution theories developed by Newman
Nernst-Planck equation, including the concentration-dependent diffu et al. [94]. The model is termed "pseudo" because its second dimension
sion and migration effects. This model provides a more comprehensive follows the electrode’s radius, rather than being a true second dimen
representation of the electrochemical processes in the battery. However, sion. This model functions similarly to the 1D model but also in
it is quite complex and computationally expensive [93]. The Newman corporates the impact of diffusion in both electrolyte and solid
P2D model, as a simplified approach to the DFN model, is widely used (electrode) phases and the Butler-Volmer kinetics for simulating elec
for physics-based electrochemical-thermal modeling. Therefore, in this trode kinetics [95–97]. Therefore, in some literature [98–103], the
study, the review approach for physics-based electrochemical-thermal Newman P2D model is also expressed as a physically based 1D elec
models focuses only on the Newman P2D model. trochemical model. It should be emphasized that, despite some studies
7
[104] claiming to employ a 3D electrochemical model by providing the broad knowledge of parameters, some of which are elusive or arduous to
governing equations in form of three dimensions, they actually simplify measure with precision, thereby increasing complexity. Moreover, the
the model during the solution process. Many variables, such as elec computational burden is substantial because of the spatial and temporal
trolyte concentration and current density, are only solved in resolutions of the models and the pseudo-second dimension for radial
one-dimensional form. As a result, there are few examples of truly diffusion within particles. In addition, accurately defining initial and
coupling a 3D electrochemical model with a 3D thermal model. There boundary conditions, particularly for complex battery geometries and
fore, this study focuses exclusively on the discussion of the widely used operating conditions, presents another layer of complexity.
P2D model. Owing to the complexity of the P2D model, independently compiling
Fig. 6. Schematically illustrates an electrode plate pair in a Li-ion and solving it can be challenging. As a result, commercially available
battery [97]. The composite electrode is composed of active materials CFD packages that integrate P2D models, such as COMSOL Multiphysics
and an electrolyte solution. This electrolyte phase consistently flows [98–116,13,117–119], are widely used for the construction of
through the negative electrode, separator, and positive electrode, physics-based electrochemical battery models. The mechanism of
whereas the solid phase is only present in the negative and positive coupling between the 1D (P2D) electrochemical model and the 3D
electrodes. The solid active material is depicted as an array of uniformly thermal model is shown in Fig. 7. The heat generation rate calculated
sized spherical particles. The separator, which is a porous polymer using the 1D model is utilized as the heat source within the 3D thermal
membrane, serves as a physical barrier that prevents the movement of model to ascertain the internal temperature distribution of the battery.
active material particles between the two electrodes. During the dis Consequently, the average temperature derived from the 3D model in
charging process, lithium migrates to the surface of the negative elec fluences the 1D electrochemical model, forming a feedback loop. This is
trode particles, where it undergoes an electrochemical reaction. This because the heat generation rate in the 1D model is sensitive to varia
reaction emits an electron and transfers lithium into the electrolyte tions in temperature-dependent electrochemical parameters, which, in
phase. The lithium ions then traverse the electrolyte solution and move turn, are affected by increasing temperatures [102,103,118,111].
from the negative electrode to the positive electrode. At the positive Although the P2D thermal coupled battery model boasts high compu
electrode, a corresponding reaction incorporates lithium into the posi tational accuracy, it involves several variables (Table 5), and the
tive solid phase. Lithium remains stored within the positive electrode simulation process requires solving several partial differential equations
particles until the battery is recharged at a later time. The Li-ion trans (PDEs) as shown in Table 4. Thus, as observed from the reviewed
port phenomena in the porous electrode and electrolyte can be described literature, only a few studies focus on the battery module with multi-cell
by charge and mass conservation laws. Charge conservation governs [109,115,119]. Most studies have focused on the thermal performances
phase potentials, φ+ and φ-, whereas mass conservation governs the of individual battery cells. In ANSYS FLUENT, the P2D model can be
phase concentrations, ce and cs, where the subscripts e and s denote the coupled with a 3D thermal model using the multiscale multidimensional
electrolyte and solid (particle) phases, respectively. The conservation (MSMD) solution method. This method is beneficial because it can
equations for the electrolyte and solid phases are listed in Table 4. In the capture battery behavior at various scales, leading to a more accurate
electrode area, the lithium conservation diffusion equation must be and detailed understanding of battery performance. For example, it can
addressed for all discretized spatial zones. The lithium conservation account for interactions between electrode particles, the effects of
equation is computed within the radial dimension of the spherical par temperature on battery performance, and the influence of various
ticles, constituting the pseudo-second dimension. Consequently, this operating conditions. However, few studies have used the P2D model for
model is frequently referred to in the literature as the Newman P2D the thermal analysis of lithium batteries in FLUENT [97,111,90]. This
model. The Butler-Volmer equation is used to couple the charge and may be because FLUENT is more suitable for solving fluid and heat
species governing equations by describing the current flux at the active transfer models than COMSOL, whereas COMSOL is more adept at
material pore-wall surface due to lithium intercalation reaction jth as a solving multiphysics problems related to battery electrochemical
function of over potential η [95]. Compared with the two types of models.
models previously introduced, the complexity inherent in solving
physics-based electrochemical models, such as the Newman P2D model, 6. Modeling of thermal runaway
arises from the intricate interplay of multiple factors. This model is
based on partial differential equations (PDEs), which encapsulate the The emphasis on safety in the context of battery design has signifi
nonlinear and interdependent nature of various physical and chemical cantly grown. In conditions of misuse, batteries can experience a state
phenomena, such as diffusion and reaction kinetics. This necessitates a known as thermal runaway. This state can be triggered by a rise in
temperature. In such scenarios, the excess release of heat could harm the
battery cell, potentially leading to fires or even explosions. It’s crucial to
understand that thermal runaway reactions are highly complex and
differ greatly based on the materials involved. The processes that trigger
thermal runaway involve mechanical, electrical, and thermal abuse
mechanisms [121]. Mitigation tactics can be implemented at different
stages - from the materials used, to the individual cell, and the system as
a whole - to ensure the comprehensive safety of energy storage systems
using LIBs. These strategies can be rooted in chemical, mechanical,
electrical, or thermal techniques, provided they accomplish any of the
following steps aimed at mitigating failure [122]:
(1) Lower the likelihood of hazardous conditions occurring;

(2) Remove hazardous conditions once they’ve manifested;
(3) Improve the battery cell’s thermal stability in the face of haz
ardous conditions;
(4) Lessen the energy discharged in the event of Thermal Runaway;
(5) Detect potential TR and issue early warnings;
(6) Control the risk of propagation and confine damage to a minimal
Fig. 6. Electrode and particle domains in the Newman model [97]. area;
8
Table 4
Governing equations of thermal coupled Newman P2D model [95,97].
Electrochemical model Governing equations
( )
Lithium conservation in the solid phase: ∂cs 1 ∂ ∂c s
= Ds r2 (28)
∂t r2 ∂r ∂r
( )
Lithium conservation in the electrolyte phase: ∂(εe ce ) ∂ ∂ c e
0
as (1 − t+ ) int ∑ side
= Deff + (j + j ) (29)
∂t ∂x e ∂ x F
( )
Charge conservation in solid phase: ∂ ∂φ ∑
σeff s − as (jint + jside ) = 0 (30)
∂x ∂x
( ) ( )
Charge conservation in the electrolyte phase: ∂ ∂φ ∂ eff ∂lnce ∑
κeff e + κ + as (jint + jside ) = 0 (31)
∂x ∂x ∂x D ∂ x
The Butler-Volmer equation: The current flux at the active material pore-wall surface due to lithium intercalation reaction as a function of over-potential η:
) )}
{ ( αa F (α F
jint = i0 exp η − c η (32)
RT RT
Where η is defined by:
∑
η = φs − φe − U − Rfilm (jint + jside ) (33)
And i0 is the exchange current density defined by:
i0 = Fkm (ce )αa (cs,max − cs,e )αa (cs,e )αc (34)
Effective properties: eff
De = De εβe (35)
eff
κ = κεβe (36)
2RTκeff 0 ( dlnf )
κeff (t+ − 1) 1 + (37)
±
D =
F dlnce
σeff = σεβs (38)
αs = 3εs /rs (39)
[ ]
Ed
Ds = Ds,ref exp − (1 ) (40)
1
R −
T Tref
[ ]
Ed
km = km,ref exp − ( ) (41)
1 1
R −
T Tref
Thermal model Governing equations
The energy conservation equation ∂T
ρCp = ∇(λ∇T) + q (42)
∂t
q = qre + qohm + qact + qex (43)
∂U
qre = jint T (44)
∂T
2RTκeff 0 ( dlnf )
qohm = σeff (∇φs )2 + κeff (∇φe )2 + (t+ − 1) 1 + ⋅∇(lnce )⋅∇φe (45)
±
F dlnce
int
qact = j (φs − φe − U) (46)
2h
qex = (Tamb − T) (47)
Th
Fig. 7. Coupling between electrochemical model and thermal model using average values for the temperature and generated heat [118].
Traditional safety tests are often costly and offer limited reproduc design for LIBs at both cell and system stages [122]. A so called
ibility. There is an immediate need for numerical methods to lower the Four-Equation Model is widely used for modeling thermal runaway re
expenses related to these safety tests, especially crucial for safety actions and simulating a battery thermal behavior under thermal abuses.
designing at a system-wide level. A dependable model can virtually This model does not take into account reactions involving lithium metal
create a safety envelope, defining a protective limit for battery systems. and combustion processes. However, the methodology introduced here
At present, Thermal runaway models are employed to inform safety could be easily adapted to include such reactions. In the model, thermal
9
Table 5 ⃒ ⃒ ⃒ ⃒
∑ ⃒dci ⃒ ⃒dSoc⃒
Battery parameters used in Newman P2D model [96,120]. q̇abuse = Hi Wi ⃒⃒ ⃒⃒ + Hec ⃒⃒ ⃒ (53)
dt dt ⃒
Parameters Negative Separator Positive Unit
electrode electrode
where Hi and Wi are the heat of reaction (J/kg) and density of reactants
Thickness ln ls lp µm (kg/m3) in the medium i(sei, ne, pe, e), respectively. The physical and
Particle radius Rsn Rsp
kinetic parameters used in the above equations are listed in Table 6.
– µm
Maximum solid Li+ cs,n,max – cs,p,max mol/m3
Concentration for
electrode 7. Parameter identification for thermal coupled battery
Initial solid Li+ cs,n,0 – cs,n,0 mol/m3 modeling
concentration for
electrode
Initial electrolyte Li+ ce,0 ce,0 ce,0 mol/m3 To evaluate the battery heat generation using the simple model and
concentration ECM, determination of the parameters, such as R0, Ri, Ci, and dUOC/dT,
Volume fraction for εn – εp – or ▽S/nF, is a key step. Various methods have been developed to
electrode
determine the internal impedance of LIBs. The parameter identification
Volume fraction for – εs – –
separator for simple models and ECMs can be discussed through the following two
Filler fraction for εf,n – εf,p – aspects: the identification of entropy coefficient and that of internal
electrode resistance and RC, respectively. For physics-based electrochemical
Reference diffusivity Ds,n – Ds,p m2/s models, parameter identification is a critical step in the development of
for electrode
thermal coupled battery models for predicting battery heat generation.
Stoi. at 0 % SOC for Stoin,0 – Stoip,0 –
electrode This process involves determining the values of various model param
Stoi. at 100 % SOC for Stoin,100 – Stoip,100 – eters, such as diffusion coefficients, reaction rates, and thermal prop
electrode erties, which are crucial for accurately capturing the physical and
Conductivity for S/m
σn – σp
chemical processes occurring within the battery. Given the diverse
electrode
Reference rate constant kn – kp mol/m2 s/
methods for battery model parameter identification, this study provides
for electrode (mol/m3) a simple review and illustrative examples of commonly used parameter
1.5 identification methods for battery thermal models.
Electrolyte diffusivity De m2/s
for electrode
Conductivity for σe S/m 7.1. Parameter identification of entropy coefficient
electrolyte
t+ factor for electrode t+ –
Nominal capacity of C Ah
The entropy coefficient of a battery cell can be evaluated using two
cell primary methods: potentiometric and calorimetric methods [120]. The
Minimum stop voltage Vmin V
Maximum stop voltage Vmax V
Table 6
Reference temperature T ref ◦
C
Universal gas constant R kJ/kg mole
Physical and kinetic parameters used for thermal abuse model [123–125].
K Parameters Description Unit
Faraday’s constant F Columb/ 1
mole Asei Decomposition frequency factor of the SEI film. s−
1
Ane Frequency factor of the reaction between negative electrode s−
and electrolyte.
1
Ape Frequency factor of the reaction between positive electrode s−
runaway reactions are put into four categories: solid electrolyte inter
and electrolyte.
face (SEI) decomposition reactions, negative electrode-electrolyte re Ae Frequency factor of electrolytic decomposition. s− 1
actions, positive electrode-electrolyte reactions, and electrolyte Ace Frequency factor of short-circuit electrochemical reaction. s− 1
decomposition reactions [123]. To monitor the progression of each kind Esei SEI-decomposition activation energy. J/
of reactions, a collective Arrhenius kinetics rate is employed [95,123]: mol
Ene Activation energy of the reaction between negative electrode J/
[ ] and electrolyte. mol
dcsei Esei msei
= − Asei exp − c (48) Epe Activation energy of the reaction between positive electrode J/
dt RT sei
and electrolyte. mol
[ ] [ ] Ee Activation energy of electrolytic decomposition. J/
dcne tsei Ene mne mol
= − Ane exp − exp − c (49)
dt tsei,ref RT ne Ece Activation energy of short-circuit electrochemical reaction. J/
mol
[ ] csei Dimensionless amount of lithium-containing meta-stable –
dα Epe mpe,1
= Ape exp − α (1 − α)mpe,2 (50) species in the SEI.
dt RT cne Dimensionless amount of lithium intercalated within the –
[ ] carbon.
dce E e me α Conversion degree of the positive electrode material. –
= − Ae exp − c (51) ce Proportion of the remaining electrolyte.
dt RT e –
msei Reaction order for csei –
In addition, in this four equation framework, the abuse reaction due mne Reaction order for cne –
to the internal short circuit can be added and governed by the lowing mpe,1 Reaction order for α –
mpe,2 Reaction order for (1- α)
kinetics rate equation [95,124]:
–
me Reaction order for ce –
[ ] tsei Dimensionless thickness of the SEI film. –
dSoc Eec
= − ISCcond ⋅Soc⋅Aec exp − (52) tsei,ref Reference SEI layer thickness. –
dt RT ISCcond The factor used to make sure that the reaction is activated –
only after the specified trigger temperature
Then the total heat generation due to the thermal runaway can be
Hi Heat of reaction. J/kg
evaluated by: Wi Density of reactants in the medium. kg/
m3
10
potentiometric method is widely used for determining the entropy co compared to other laboratory equipment. Consequently, the time and
efficient of a battery. It involves measuring the OCV of a battery at cost associated with EIS present significant drawbacks, making this
various temperatures while maintaining a constant SOC. To conduct a approach less practical for rapid, cost-effective, and large-scale param
potentiometric measurement, the battery is initially charged or dis eter identification. Current-pulse-based characterization techniques,
charged to a certain SOC and subsequently left to rest until it reaches which employ constant current pulses of a fixed duration to examine the
thermal equilibrium. The OCV is then measured at various tempera transient response of a cell, can yield satisfactory results in a shorter
tures. From these measurements, the change in OCV with temperature time and at a lower cost than EIS. Consequently, the simplicity of
(dUOC/dT) can be calculated, providing the entropy coefficient. This pulse-based characterization has gained widespread popularity [71]. In
method provides a direct and relatively simple way to measure the en a study by Jiaqiang et al. [42], the detailed process of an HPPC test for
tropy coefficient and can be applied to various types of batteries. the identification of battery internal resistance was presented. In this
Conversely, the calorimetric method is another technique used to study, the HPPC test begins with a charged device up to 4.2 V of the cell
determine the entropy coefficient of a battery. It involves directly voltage. Following a one-hour rest, an HPPC profile with a 30-second
measuring the heat produced or absorbed by the battery during its discharge pulse, 40-second rest, and 10-second regenerate pulse is per
operation, which is related to the entropy change in the battery. Clearly, formed. Subsequently, it is followed by a discharge to the next 10 %
this method cannot be used for numerical analysis of battery heat gen increment of the capacity at the C/3 discharge rate. Finally, it undergoes
eration. Therefore, the potentiometric method is commonly used to another one-hour rest. This sequence is repeated until the final profile
identify the entropy coefficient for modeling battery thermal models, near 10 % SOC is achieved. The test should terminate at a cell voltage of
even with its highly time-consuming process [126]. Based on the liter 2.75 V, which is recommended by the manufacturer. Fig. 9(a) shows the
ature reviewed in this study, not all studies provide an illustration of the test process, and Fig. 9(b) shows one section of discharge pulses of the
detailed experimental procedures for the potentiometric method. Jia HPPC test at the SOC of 100 % under 0 ◦ C. The instant voltage decrease
qiang et al. [42] conducted an experiment to determine the entropy from V1 to V2 is a result of the ohmic resistance, whereas the decrease
coefficient. The experimental procedure illustrated in Fig. 8 involves six from V2 to V3 is caused by the cell polarization resistance. Therefore, R0
identical 18,650 cells each at a SOC: 0 %, 20 %, 40 %, 60 %, 80 %, and at an SOC of 100 % and 0 ◦ C can be acquired using Eq. (49). Finally, the
100 %. These cells are initially placed in a temperature chamber internal resistance can be obtained using a nonlinear curve-fitting
maintained at 40 ◦ C. After a rest period of 2 h, the cell voltages (V1) are method with multiple variables as functions of the battery SOC and
measured and logged. Subsequently, the temperature in the chamber is temperature.
adjusted to 10 ◦ C, and after another 2 h of resting period, the voltages
V1 − V2 V2 − V3 V1 − V3
(V2) of the cells are recorded again. The parameter dE/dT, specific to R0 = + = (55)
I I I
each SOC, is computed from these voltage measurements using Eq. The
dUoc/dT can then be fitted as a function of the SOC using a nonlinear
polynomial method. 7.3. Parameter identification of RC network
dUoc V1 − V2
= (54) Similar to the parameter identification method for the internal
dT T1 − T2
resistance, the experimental data required for the parameter identifi
cation of the RC network can also be obtained using the HPPC and GITT
7.2. Parameter identification of internal resistance methods. The difference lies in the adaptation of the two parameters in
the RC network, Ri and Ci, which typically require the use of an iterative
Several experimental methods can be employed to identify the in parameter identification algorithm (IPIA). GITT is performed in the
ternal resistance and capacitance of a battery, such as current pulse test, study of Felix-A. LeBel [71]. The process employs a constant current
hybrid pulse power characterization (HPPC) test, electrochemical titration of 1C (3.5 A) for 30 s, inducing an SOC change of 0.8 %. This is
impedance spectroscopy (EIS), galvanostatic intermittent titration followed by a 600-second rest period to allow the electrochemical
technique (GITT). Two predominant methods exist in this field: elec properties to regain equilibrium. If the cell voltage is greater than or
trochemical impedance spectroscopy (EIS) and current-pulse-based equal to 2.4 V at the end of the rest period, further titrations are per
characterization [71]. EIS, as a group of nondestructive techniques, formed, thus conducting GITT across the entire SOC spectrum. If the
measures the reaction rate of an electrochemical system by applying a voltage drops below 1.4 V during titration, the process is halted to
sinusoidal current or voltage across a range of frequencies. However, it prevent potential damage to the cell. The GITT sequence was performed
typically requires high-fidelity, expensive galvanostats or potentiostats at constant temperatures at − 20 ◦ C, 0 ◦ C, 20 ◦ C, 40 ◦ C, and 60 ◦ C. In the
Fig. 8. Test process for the identification of entropy coefficient [42].
11
Fig. 9. HPPC test for the identification of R0 [42].
study, the IPIA is introduced. The transient forms of Eq. (24) for the DP
model can be represented as
∑
n
Ub (t) = UOC (t) − Ib (t)R0 (SOC, T) − Ui (t) (56)
i=1,2
Ui (t) = Ri (SOC, T)I(t)e− t/Ri (SOC,T)Ci (SOC,T)

(57)
∑
n
Up (t) = c + ai e− t/τi
(58)
i=1,2
The time constant τi=RiCi of each RC network can then be rewritten

as Eq. (52), where constant values are regrouped in constant c. Subse
quently, the identification of the parameters can be obtained by the
nonlinear least square method curve-fitting on the voltage Up with the
objective function given in Eq. (52), where ai is a fitting coefficient. Once
Fig. 10. Representation of the iterative parameter identification algorithm
the coefficients are fitted, the values of Ri and Ci are obtained using Eq.
(IPIA) [71].
(53) as [89]:
Ri =
ai
; Ci =
1
(59) were performed by Shi et al. [78] and Saqli et al. [79]. The detailed
Ip ai parameter identification methods used in the reviewed studies utilizing
simple models and ECMs are summarized in Table 7.
Fig. 10 illustrates the basic concept of the IPIA. Initially, the algo
rithm executes a first-order parameter identification, and the results a1*
and t1* are then employed as the initial values for a1 and t1 in the first- 7.4. Parameter identification of physics-based electrochemical model
order terms of a second-order equation. The second-order parameters a2
and t2 are assigned starting values that align with the magnitude order The reliability of the Newman P2D model is significantly influenced
where the solution is expected to converge relative to a1 and t1. The by the accuracy of its parameters, which are tailored to the unique
parameter identification results are presented in Fig. 11. Similar studies design and chemistry of each cell. Hence, the parameters from the one-
cell design are not universally applicable. This makes the process of
12
Fig. 11. Identified 1st-order and 2nd-order RC (DP) model parameters [71].
finding an appropriate set of parameters for a specific battery one of the specification sheets. As a common practice within the modeling com
primary challenges in battery modeling. Some researchers may attempt munity, parameter sets are often sourced from literature, albeit without
to fit the model for alignment with the experimentally measured voltage necessarily tracing their origins [127]. Similarly, it can be observed in
[114]. However, because of the complexity of the model and the this study that almost all the reviewed work did not explicitly state their
extensive number of parameters, this method is often not viable without methods for parameter identification.
a well-informed initial estimate of the parameters. Using parameters In this case, owing to the complexity of the parameter identification
sourced from the literature is another option; however, inconsistencies method for the P2D model, a detailed explanation is not provided. In a
can occur if the parameters are collected from various sources, poten study by Chen et al. [127], OCV, lithium stoichiometry, activation en
tially leading to inaccurate predictions. Therefore, the experimental ergy, and exchange current coefficient were obtained using GITT and
measurement of these parameters could offer a more robust method, EIS measurements. The apparent diffusion coefficients were determined
potentially leading to more accurate model predictions [127]. The by applying the Sand equation to the voltage transient that occurs
Newman P2D model requires more than 30 parameters to accurately during the current pulse test. Li et al. [128] identified the diffusion co
depict the physical, chemical, and electrochemical properties of a cell, efficient, surface reaction rate, and exchange current density using
enabling the simulation of its behavior. The precise value depends on the analytical solutions from potentiostatic intermittent titration technique
definition of the model. Determining these parameter values is a com (PITT) measurements. The aforementioned studies can provide a refer
plex and time-consuming task, necessitating various characterization ence for the estimation and identification of parameters in P2D models.
and analytical techniques. This task is even more demanding in com To obtain more accurate model parameters, a profound understanding
mercial cell chemistry because manufacturers do not typically disclose of the electrochemical principles of batteries and more precise experi
the parameter, composition, or component information in their mental equipment and rational experimental methods, are required.
13
Table 7
Parameter identification methods used in the studies utilizing simple models and ECMs.
Author Test Capacity Amplitude Pulse duration Rest duration ΔSOC Temperature [ ◦ C] No. of Thermal model
method [Ah] [s] [s] RC
Han et al. [41] – 94 0.33C – 3h 5% 25 0 3D CFD

E et al. [42] HPPC 2.6 0.33C 30s 40s 10 % 0, 15, 30, 45 0 3D CFD
Tan et al. [47] Pulse test 3.2 0.3C 600 7200s – 20, 30, 40 0 3D CFD
Shen et al. [48] HPPC 8 3C, 5C, 7C, 9C – – – 30 0 3D CFD
Kang et al. [53] Pulse test 3.3 1C 180 3600 5% 26 0 3D CFD
Wang et al. [63] HPPC 35 1C 10 40 5% 25 2 –
Estevez et al. Pulse test 59.3 1C 30 – 10 % 0, 25, 40 1 3d Lumped
[68]
Zhu et al. [69] Pulse test 2.5 1C ~ 10C 30 7200 – − 20 ~ 45 1 1D-Lumped
He et al. [74] HPPC 43 – – 3600 5% 0, 10, 25, 45 2 1D-Lumped/
CFD
Hou et al. [75] HPPC 22 0.5C 20 40 10 % 20 2 1D-Lumped
Wang et al. [76] Pulse test 2.5 1C, 2C – 7200 – 0 ~ 45 2 1D-Lumped
Saqli et al. [79] HPPC 12 1C 10 40 10 % − 10, 0, 15, 25, 35, 2 3D CFD
45
Behi et al. [80] HPPC 23 8C 10 10 – 25 2 3D CFD
Shi et al. [78,88] HPPC 22 0.5C, 1C, 2C, – 30 5% 5, 25, 45 2 1D-Lumped
3C
Damay et al. GITT – 0.2C ~ 2C – 1800 2.50 % 5 ~ 45 1 1D-Lumped
[87]
8. Conclusions methodology introduced here could be easily adapted to include

such reactions.
The thermal management and the accurate prediction of heat gen (5) The key parameters for thermally coupled battery modeling were
eration in Lithium-ion batteries remain to be significant challenges in identified. Considering the wide range of methodologies for
battery technology, thereby impacting the safety, performance, and identifying parameters in battery models, this paper offers an
lifespan of batteries in applications such as electric vehicles and energy overview and example-based explanation of several frequently
storage systems. This study provides a comprehensive analysis of employed techniques for parameter identification within the
various thermal coupled battery models, including simple battery thermal models of batteries.
models, ECM, and physics-based electrochemical models. Moreover, (6) This study presents some limitations that should be addressed.
parameter identification methods for these models have been discussed. For example, thermal models within coupled models have not
The general conclusions of this review are as follows: been extensively elaborated. Additionally, less commonly used
battery thermal models, such as the NTGK model [129–134],
(1) Simple battery models cannot fully reflect the dynamic and were not considered.
electrochemical characteristics of a battery and are typically
coupled with three-dimensional CFD thermal models to design, Declaration of Competing Interest
evaluate, optimize, and improve battery thermal management
systems. These studies are often not particularly sensitive to the The authors declare that they have no known competing financial
inherent properties of batteries. interests or personal relationships that could have appeared to influence
(2) ECM, owing to its intricate structure, offers a broader depiction of the work reported in this paper.
battery performance, integrating numerous resistive and capaci
tive elements to better portray the dynamic behavior. The Rint Data availability
model simplifies this process by employing a singular resistance
element. This complexity generally enhances the precision of Data will be made available on request.
ECM when coupled with a CFD model to forecast the battery’s
response under diverse operating conditions. Even with regard to
thermal predictions, the elaborate portrayal of internal battery Acknowledgements
procedures using ECM could lead to more precise outcomes.
(3) The Newman P2D model is widely used in various physics-based This work was supported by the 2020 Yeungnam University
electrochemical methods. Although this model has high compu Research Grant.
tational accuracy, its complexity and numerous parameters and
variables make it computationally intensive to solve. Hence, this References
model is typically used only to analyze the performance of indi
vidual battery units. Simplifying the model while ensuring [1] J. Jaguemont, J.V. Mierlo, A comprehensive review of future thermal
computational accuracy and applying it to the simulation of management systems for battery-electrified vehicles, J. Energy Storage 31
(2020), 101551.
large-scale battery pack thermal management systems is a chal [2] E. Alptekin, Emission, injection and combustion characteristics of biodiesel and
lenging task. oxygenated fuel blends in a common rail diesel engine, Energy 119 (2017) 44–52.
(4) In addition, a so called Four-equation thermal abuse model is also [3] P.R. Tete, M.M. Gupta, S.S. Joshi, Developments in battery thermal management
systems for electric vehicles: a technical review, J. Energy Storage 35 (2021),
reviewed. . In the model, thermal runaway reactions are put into 102255.
four categories: solid electrolyte interface (SEI) decomposition [4] A.K. Thakur, R. Prabakaran, M.R. Elkadeem, S.W. Sharshir, M. Arıcı, C. Wang,
reactions, negative electrode-electrolyte reactions, positive W. Zhao, J.Y. Hwang, R. Saidur, A state of art review and future viewpoint on
advance cooling techniques for Lithium–ion battery system of electric vehicles,
electrode-electrolyte reactions, and electrolyte decomposition J. Energy Storage 32 (2020), 101771.
reactions. The model does not take into account reactions [5] Z. Shen, S. Chen, X. Liu, B. Chen, A review on thermal management performance
involving lithium metal and combustion processes. However, the enhancement of phase change materials for vehicle lithium-ion batteries, Renew.
Sustain. Energy Rev. 148 (2021), 111301.
14
[6] J. Kim, J. Oh, H. Lee, Review on battery thermal management system for electric [36] H. Behia, D. Karimi, M. Behi, M. Ghanbarpour, J. Jaguemont, M.A. Sokkeha, F.
vehicles, Appl. Therm. Eng. 149 (2019) 192–212. H. Gandoman, M. Berecibar, J.V. Mierlo, A new concept of thermal management
[7] V. Mali, R. Saxena, K. Kumar, A. Kalam, B. Tripathi, Renew. Sustain. Energy Rev. system in Li-ion battery using air cooling and heat pipe for electric vehicles, Appl.
151 (2021), 111611. Therm. Eng. 174 (2020), 115280.
[8] Q.L. Yue a, C.X. He, M.C. Wu, T.S. Zhao, Advances in thermal management [37] M. Akbarzadeh, T. Kalogiannis, J. Jaguemont, L. Jin, H. Behi, D. Karimi,
systems for next-generation power batteries, Int. J. Heat Mass Transf. 181 (2021), H. Beheshti, J.V. Mierlo, M. Berecibar, A comparative study between air cooling
121853. and liquid cooling thermal management systems for a high-energy lithium-ion
[9] D.K. Sharma, A. Prabhakar, A review on air cooled and air centric hybrid thermal battery module, Appl. Therm. Eng. 198 (2021), 117503.
management techniques for Li-ion battery packs in electric vehicles, J. Energy [38] H. Liu, S. Ahmad, Y. Shi, J. Zhao, A parametric study of a hybrid battery thermal
Storage 41 (2021), 102885. management system that couples PCM/copper foam composite with helical liquid
[10] L. Yang, F. Zhou, L. Sun, S. Wang, Thermal management of lithium-ion batteries channel cooling, Energy 231 (2021), 120869.
with nanofluids and nano-phase change materials: a review, J. Power Sources 539 [39] H. Wang, T. Tao, J. Xu, X. Mei, X. Liu, P. Gou, Cooling capacity of a novel
(2022), 231605. modular liquid-cooled battery thermal management system for cylindrical
[11] A.A. Hakeem Akinlabi, D. Solyali, Configuration, design, and optimization of air- lithium ion batteries, Appl. Therm. Eng. 178 (2020), 115591.
cooled battery thermal management system for electric vehicles: a review, [40] S. Lee, U. Han, H. Lee, Development of a hybrid battery thermal management
Renew. Sustain. Energy Rev. 125 (2020), 109815. system coupled with phase change material under fast charging conditions,
[12] Wang Zichen, Du Changqing, A comprehensive review on thermal management Energy Conv. Manag. 268 (2022), 116015.
systems for power lithium-ion batteries, Renew. Sustain. Energy Rev. 139 (2021), [41] U. Han, H. Kang, J. Song, J. Oh, H. Lee, Development of dynamic battery thermal
110685. model integrated with driving cycles for EV applications, Energy Conv. Manag.
[13] Z. Chen, Y. Qin, Z. Dong, J. Zheng, Y. Liu, Numerical study on the heat generation 250 (2021), 114882.
and thermal control of lithium-ion battery, Appl. Therm. Eng. 221 (2023), [42] J. E, M. Yue, J. Chen, H. Zhu, Y. Deng, Y. Zhu, F. Zhang, M. Wen, B. Zhang,
119852. S. Kang, Effects of the different air cooling strategies on cooling performance of a
[14] S. Abada, G. Marlair, A. Lecocq, M. Petit, V. Sauvant-Moynot, F. Huet, Safety lithium-ion battery module with baffle, Appl. Therm. Eng. 144 (2018) 231–241.
focused modeling of lithium-ion batteries: a review, J. Power Sources 306 (2016) [43] J. Cao, M. Luo, X. Fang, Z. Ling, Z. Zhang, Liquid cooling with phase change
178–192. materials for cylindrical Li-ion batteries: an experimental and numerical study,
[15] V.G. Choudharia, A.S. Dhoblea, T.M. Sathe, A review on effect of heat generation Energy 191 (2020), 116565.
and various thermal management systems for lithium ion battery used for electric [44] L.H. Sawa, H.M. Poona, H.S. Thiama, Z. Cai, W.T. Chong, N.A. Pambudi, Y.
vehicle, J. Energy Storage 32 (2020), 101729. J. King, Novel thermal management system using mist cooling for lithium-ion
[16] X. Zhang, Z. Li, L. Luo, Y. Fan, Z. Du, A review on thermal management of battery packs, Appl. Energy 223 (2018) 146–158.
lithium-ion batteries for electric vehicles, Energy 238 (2022), 121652. [45] Z. Tang, S. Wang, Z. Liu, J. Cheng, Numerical analysis of temperature uniformity
[17] Y. Zeng, D. Chalise, S.D. Lubner, S. Kaur, R.S. Prasher, A review of thermal of a liquid cooling battery module composed of heat-conducting blocks with
physics and management inside lithium-ion batteries for high energy density and gradient contact surface angles, Appl. Therm. Eng. 178 (2020), 115509.
fast charging, Energy Storage Mater 41 (2021) 264–288. [46] Kausthubharam, P.K. Koorata, N. Chandrasekaran, Numerical investigation of
[18] G. Zhao, X. Wang, M. Negnevitsky, H. Zhang, A review of air-cooling battery cooling performance of a novel air-cooled thermal management system for
thermal management systems for electric and hybrid electric vehicles, J. Power cylindrical Li-ion battery module, Appl. Therm. Eng. 193 (2021), 116961.
Sources 501 (2021), 230001. [47] X. Tan, P. Lyu, Y. Fan, J. Rao, K. Ouyang, Numerical investigation of the direct
[19] Y. Denga, C. Feng, J. E, H. Zhua, J. Chen, M. Wen, H. Yin, Effects of different liquid cooling of a fast-charging lithium-ion battery pack in hydrofluoroether,
coolants and cooling strategies on the cooling performance of the power lithium Appl. Therm. Eng. 196 (2021), 117279.
ion battery system: a review, Appl. Therm. Eng. 142 (2018) 10–29. [48] L. Sheng, L. Su, H. Zhang, K. Li, Y. Fang, W. Ye, Y. Fang, Numerical investigation
[20] C. Roe, X. Feng, G. White, R. Li, H. Wang, X. Rui, C. Li, F. Zhang, V.r Null, on a lithium ion battery thermal management utilizing a serpentine-channel
M. Parkes, Y. Patel, Y. Wang, H. Wang, M. Ouyang, G. Offer, B. Wu, Immersion liquid cooling plate exchanger, Int. J. Heat Mass Transf. 141 (2019) 658–668.
cooling for lithium-ion batteries – A review, J. Power Sources 525 (2022), [49] G. Lian, J. Li, J. He, J. Tian, X. Chen, L. Chen, Numerical investigation on a
231094. unitization-based thermal management for cylindrical lithium-ion batteries,
[21] M. Bernagozzi, A. Georgoulas, N. Mich´e, M. Marengo, Heat pipes in battery Energy Rep 8 (2022) 4608–4621.
thermal management systems for electric vehicles: a critical review, Appl. Therm. [50] Z. An, X. Chen, L. Zhao, Z. Gao, Numerical investigation on integrated thermal
Eng. 219 (2023), 119495. management for a lithiumion battery module with a composite phase change
[22] D.M. Weragoda, G. Tian, A. Burkitbayev, K. Lo, T. Zhang, A comprehensive material and liquid cooling, Appl. Therm. Eng. 163 (2019), 114345.
review on heat pipe based battery thermal management systems, Appl. Therm. [51] P. Saechan, I. Dhuchakallay, Numerical study on the air-cooled thermal
Eng. 224 (2023), 120070. management of Lithium-ion battery pack for electrical vehicles, Energy Rep 8
[23] G. Murali, G.S.N. Sravya, J. Jaya, V.N. Vamsi, A review on hybrid thermal (2022) 1264–1270.
management of battery packs and it’s cooling performance by enhanced PCM, [52] D.S. Jang, S. Yun, S.H. Hong, W. Cho, Y. Kim, Performance characteristics of a
Renew. Sustain. Energy Rev. 150 (2021), 111513. novel heat pipe-assisted liquid cooling system for the thermal management of
[24] J. Luo, D. Zou, Y. Wang, S. Wang, L. Huang, Battery thermal management systems lithium-ion batteries, Energy Conv. Manag. 251 (2022), 115001.
(BTMs) based on phase change material (PCM): a comprehensive review, Chem. [53] T. Kang, P.Yeon. Lee, S. Kwon, K. Yoo, J. Kim, Regional resistance-based spatial
Eng. J. 430 (2022), 132741. thermal model for checking non-uniformed temperature distribution and
[25] A. Wazeer, A. Das, C. Abeykoon, A. Sinha, A. Karmakar, Phase change materials evolution of pouch type lithium-ion batteries, Appl. Therm. Eng. 192 (2021),
for battery thermal management of electric and hybrid vehicles: a review, Energy 116936.
Nexus 7 (2022), 100131. [54] M. Shen, Q. Gao, Structure design and effect analysis on refrigerant cooling
[26] Z. Yu, J. Zhang, W. Pan, A review of battery thermal management systems about enhancement of battery thermal management system for electric vehicles,
heat pipe and phase change materials, J. Energy Storage 62 (2023), 106827. J. Energy Storage 32 (2020), 101940.
[27] J.R. Patel, M.K. Rathod, Recent developments in the passive and hybrid thermal [55] Y. Huang, S. Wang, Y. Lua, R. Huang, X. Yu, Study on a liquid cooled battery
management techniques of lithium-ion batteries, J. Power Sources 480 (2020), thermal management system pertaining to the transient regime, Appl. Therm.
228820. Eng. 180 (2020), 115793.
[28] A. Fotouhi, D.J. Auger, K. Propp, S. Longo, M. Wild, A review on electric vehicle [56] Y. Chung, M.Soo. Kim, Thermal analysis and pack level design of battery thermal
battery modelling: from Lithium-ion toward Lithium-Sulphur, Renew. Sustain. management system with liquid cooling for electric vehicles, Energy Conv.
Energy Rev. 56 (2016) 1008–1021. Manag. 196 (2019) 105–116.
[29] C. Zhang, K. Li, S. Mcloone, Z. Yang, Battery Modelling Methods for Electric [57] W. Yanga, F. Zhou, H. Zhoua, Q. Wang, J. Kong, Thermal performance of
Vehicles - A Review, in: Proceedings of the European Control Conference (ECC), cylindrical lithium-ion battery thermal management system integrated with mini-
Strasbourg, France, 2014, pp. 2673–2678. channel liquid cooling and air cooling, Appl. Therm. Eng. 175 (2020), 115331.
[30] B. Lamrani, B...E. Lebrouhi, Y. Khattari, T. Kousksou, A simplified thermal model [58] S.Ho. Hong, D.S. Jang, S. Park, S. Yun, Y. Kim, Thermal performance of direct
for a lithium-ion battery pack with phase change material thermal management two-phase refrigerant cooling for lithium-ion batteries in electric vehicles, Appl.
system, J. Energy Storage 44 (2021), 103377. Therm. Eng. 173 (2020), 115213.
[31] D. Bernardi, E. Pawlikowski, J. Newman, A general energy balance for battery [59] Y. Wang, Y. Yu, Z. Jing, C. Wang, G. Zhou, W. Zhao, Thermal performance of
systems, J. Electrochem. Soc. 132 (1985) 5–12. lithium-ion batteries applying forced air cooling with an improved aluminum
[32] S. Wang, S. Ji, Y. Zhu, A comparative study of cooling schemes for laminated foam heat sink design, Int. J. Heat Mass Transf. 167 (2021), 120827.
lithium-ion batteries, Appl. Therm. Eng. 182 (2021), 116040. [60] A.R. Sharma, C.S. Sai, A. Kumar, R.V.J. Reddy, D. Danyharsha, R. Jilte, Three-
[33] W. Shi, J. Zheng, J. Xiao, X. Chen, B.J. Polzin, J. Zhang, The effect of entropy and dimensional CFD study on heat dissipation in cylindrical lithium-ion battery
enthalpy changes on the thermal behavior of li-mn-rich layered composite module, Mater. Today 46 (2021) 10964–10968.
cathode materials, J. Electrochem. Soc. (163–3) (2016) A571–A577. [61] F. Li, F. Lan, J. Chen, Dynamic thermal characteristics of heat pipe via segmented
[34] M. Akbarzadeh, J. Jaguemont, T. Kalogiannis, D. Karimi, J. He, L. Jin, P. Xie, J. thermal resistance model for electric vehicle battery cooling, J. Power Sources
V. Mierlo, M. Berecibar, A novel liquid cooling plate concept for thermal 321 (2016) 57–70.
management of lithium-ion batteries in electric vehicles, Energy Conv. Manag. [62] M. Baši, D. Vukadinović, V. Višnjić, I. Rakić, Dynamic equivalent circuit models of
231 (2021), 113862. lead-acid batteries - a performance comparison, IFAC PapersOnLine 55-4 (2022)
[35] S. Chen, X. Peng, N. Bao, A. Garg, A comprehensive analysis and optimization 189–194.
process for an integrated liquid cooling plate for a prismatic lithium-ion battery
module, Appl. Therm. Eng. 156 (2019) 324–339.
15
[63] Q. Wang, J. Wang, P. Zhao, J. Kang, F. Yan, C. Du, Correlation between the model [91] C. Hong, H. Cho, D. Hong, S. Oh, Y. Kim, An improved thermal single particle
accuracy and model-based SOC estimation, Electrochim. Acta 228 (2017) model and parameter estimation for high-capacity battery cell, Electrochim. Acta
146–159. 439 (2023), 141638.
[64] K.S. Hariharan, A coupled nonlinear equivalent circuit - Thermal model for [92] N. Baba, H. Yoshida, M. Nagaoka, C. Okuda, S. Kawauchi, Numerical simulation
lithium ion cells, J. Power Sources 227 (2013) 171–176. of thermal behavior of lithium-ion secondary batteries using the enhanced single
[65] W. Cao, Y. Qiu, P. Peng, F. Jiang, A full-scale electrical-thermal-fluidic coupling particle model, J. Power Sources 252 (2014) 214–228.
model for li-ion battery energy storage systems, Appl. Therm. Eng. 185 (2021), [93] F.B. Planella, M. Sheikh, W.D. Widanage, Systematic derivation and validation of
116360. a re duce d thermal-electrochemical model for lithium-ion batteries using
[66] L. Calearo, A. Thingvad, C. Ziras, M. Marinelli, A methodology to model and asymptotic methods, Electrochim. Acta 388 (2021), 138524.
validate electro-thermal-aging dynamics of electric vehicle battery packs, [94] M. Doyle, T.F. Fuller, J. Newman, Modeling of galvanostatic charge and discharge
J. Energy Storage 55 (2022), 105538. of the lithium/polymer/insertion cell, J. Electrochem. Soc. 140-6 (1993)
[67] Q. Wang, Y. He, J. Shen, Z. Ma, G. Zhong, A unified modeling framework for 1526–1533.
lithium-ion batteries: an artificial neural network based thermal coupled [95] Ansys Fluent Theory Guide.
equivalent circuit model approach, Energy 138 (2017) 118–132. [96] B. Balagopal, M. Chow, Effect of anode conductivity degradation on the thevenin
[68] M.A.P. Estevez, S. Calligaro, O. Bottesi, C. Caligiuri, M. Renzi, An electro-thermal circuit model of lithium ion batteries, in: Proceedings of the IECON - 42nd Annual
model and its electrical parameters estimation procedure in a lithium-ion battery Conference of the IEEE Industrial Electronics Society, Florence, Italy, 2016,
cell, Energy 234 (2021), 121296. pp. 2028–2033.
[69] J. Zhua, M. Knapp, M.S.D. Darma, Q. Fang, X. Wang, H. Dai, X. Wei, [97] S. Panchal, M. Mathew, R. Fraser, M. Fowler, Electrochemical thermal modeling
H. Ehrenberg, An improved electro-thermal battery model complemented by and experimental measurements of 18650 cylindrical lithium-ion battery during
current dependent parameters for vehicular low temperature application, Appl. discharge cycle for an EV, Appl. Therm. Eng. 135 (2018) 123–132.
Energy 248 (2019) 149–161. [98] M. Al-Zareer, I. Dincer, M.A. Rosen, A novel approach for performance
[70] M. Seo, Y. Song, J. Kim, S.W. Paek, G. Kim, S.W. Kim, Innovative lumped-battery improvement of liquid to vapor based battery cooling systems, Energy Conv.
model for state of charge estimation of lithium-ion batteries under various Manag 187 (2019) 191–204.
ambient temperatures, Energy 226 (2021), 120301. [99] F. Bahiraei, M. Ghalkhani, A. Fartaj, G. Nazri, A pseudo 3D electrochemical-
[71] F. LeBel, P. Messier, A. Sari, J.P.F. Troṽao, Lithium-ion cell equivalent circuit thermal modeling and analysis of a lithium-ion battery for electric vehicle
model identification by galvanostatic intermittent titration technique, J. Energy thermal management applications, Appl. Therm. Eng. 125 (2017) 904–918.
Storage 54 (2022), 105303. [100] M. Xu, Z. Zhang, X. Wang, L. Jia, L. Yang, A pseudo three-dimensional
[72] Z. Liu, X. Liu, H. Meng, L. Guo, Z. Zhang, Numerical analysis of the thermal electrochemicalethermal model of a prismatic LiFePO4 battery during discharge
performance of a liquid cooling battery module based on the gradient ratio flow process, Energy 80 (2015) 303–317.
velocity and gradient increment tube diameter, Int. J. Heat Mass Transf. 175 [101] E. Hosseinzadeh, R. Genieser, D. Worwood, A. Barai, J. Marco, P. Jennings,
(2021), 121338. A systematic approach for electrochemical-thermal modelling of a large format
[73] Z. Xu, C. Zhang, B. Sun, S. Liu, The electric-thermal coupling simulation and state lithium-ion battery for electric vehicle application, J. Power Sources 382 (2018)
estimation of lithium-ion battery, J. Energy Storage 58 (2023), 106431. 77–94.
[74] J. He, M.S. Hosen, R. Youssef, T. Kalogiannis, J.V. Mierlo, M. Berecibar, A lumped [102] S.D. Chitta, C. Akkaldevi, J. Jaidi, S. Panchal, M. Fowler, R. Fraser, Comparison of
electro-thermal model for a battery module with a novel hybrid cooling system, lumped and 1D electrochemical models for prismatic 20Ah LiFePO4 battery
Appl. Therm. Eng. 221 (2023), 119874. sandwiched between minichannel cold-plates, Appl. Therm. Eng. 199 (2021),
[75] G. Hou, X. Liu, W. He, C. Wang, J. Zhang, X. Zeng, Z. Li, D. Shao, An equivalent 117586.
circuit model for battery thermal management system using phase change [103] H. Li, C. Liu, A. Saini, Y. Wang, H. Jiang, T. Yang, L. Chen, C. Pan, H. Shen,
material and liquid cooling coupling, J. Energy Storage 55 (2022), 105834. Coupling multi-physics simulation and response surface methodology for the
[76] N. Wang, A. Chen, W. Zhao, R. Zhu, B. Duan, An online temperature estimation thermal optimization of ternary prismatic lithium-ion battery, J. Power Sources
for cylindrical lithium-ion batteries based on simplified distribution electrical- 438 (2019), 226974.
thermal model, J. Energy Storage 55 (2022), 105326. [104] H. Li, A. Saini, C. Liu, J. Yang, Y. Wang, T. Yang, C. Pan, L. Chen, H. Jiang,
[77] K. Kirad, M. Chaudhari, Design of cell spacing in lithium-ion battery module for Electrochemical and thermal characteristics of prismatic lithium-ion battery
improvement in cooling performance of the battery thermal management system, based on a three-dimensional electrochemical-thermal coupled model, J. Energy
J. Power Sources 481 (2021), 229016. Storage 42 (2021), 102976.
[78] H. Shi, S. Wang, C. Fernandez, C. Yu, W. Xu, B.E. Dablu, L. Wang, Improved [105] M. Mastalia, E. Foreman, A. Modjtahedi, E. Samadani, A. Amirfazli, S. Farhad, R.
multi-time scale lumped thermoelectric coupling modeling and parameter A. Fraser, M. Fowler, Electrochemical-thermal modeling and experimental
dispersion evaluation of lithium-ion batteries, Appl. Energy 324 (2022), 119789. validation of commercial graphite/LiFePO4 pouch lithium-ion batteries, Int. J.
[79] K. Saqli, H. Bouchareb, N.K. M’sirdi, M.O. Bentaie, Lithium-ion battery electro- Therm. Sci. 129 (2018) 218–230.
thermal modelling and internal states co-estimation for electric vehicles, [106] X. Zhang, P. Li, B. Huang, H. Zhang, Numerical investigation on the thermal
J. Energy Storage 63 (2023), 107072. behavior of cylindrical lithium-ion batteries based on the electrochemical-thermal
[80] H. Behi, D. Karimi, J. Jaguemont, F.H. Gandoman, T. Kalogiannis, M. Berecibar, coupling model, Int. J. Heat Mass Transf. 199 (2022), 123449.
J.V. Mierlo, Novel thermal management methods to improve the performance of [107] A. Greco, X. Jiang, A coupled thermal and electrochemical study of lithium-ion
the Li-ion batteries in high discharge current applications, Energy 224 (2021), battery cooled by paraffin/porous-graphite-matrix composite, J. Power Sources
120165. 315 (2016) 127–139.
[81] C. Zhang, W. Allafi, Q. Dinh, P. Ascencio, J. Marco, Online estimation of battery [108] J. Chiew, C.S. Chinb, W.D. Toh, Z. Gao, J. Jia, C.Z. Zhang, A pseudo three-
equivalent circuit model parameters and state of charge using decoupled least dimensional electrochemical-thermal model of a cylindrical LiFePO4/graphite
squares technique, Energy 142 (2018) 678–688. battery, 147 (2019) 450–463.
[82] Y. Li, Z. Zhou, W.T. Wu, Three-dimensional thermal modeling of Li-ion battery [109] M. Guo, G. Kim, R.E. White, A three-dimensional multi-physics model for a Li-ion
cell and 50V Li-ion battery pack cooled by mini-channel cold plate, Appl. Therm. battery, J. Power Sources 240 (2013) 80–94.
Eng. 147 (2019) 829–840. [110] Y. Jiang, L. Zhang, G. Offer, H. Wang, A user-friendly lithium battery simulator
[83] C. Liu, W. Liu, L. Wang, G. Hu, L. Ma, B. Ren, A new method of modeling and state based on open-source CFD, Digital Chem. Eng. 5 (2022), 100055.
of charge estimation of the battery, J. Power Sources 320 (2016) 1–12. [111] P.G. Zadeh, E. Gholamalizadeh, Y. Wang, J.D. Chung, Electrochemical modeling
[84] X. Lin, H.E. Perez, S. Mohan, J.B. Siegel, A.G. Stefanopoulou, Y. Ding, M. of a thermal management system for cylindrical lithium-ion battery pack
P. Castanier, A lumped-parameter electro-thermal model for cylindrical batteries, considering battery capacity fade, Case Stud, Therm. Eng. 32 (2022), 101878.
J. Power Sources 257 (2014) 1–11. [112] S. Panchala, M. Mathew, R. Fraser, M. Fowler, Electrochemical thermal modeling
[85] L.H. Saw, Y. Ye, A.A.O. Tay, Electro-thermal characterization of Lithium Iron and experimental measurements of 18650 cylindrical lithium-ion battery during
Phosphate cell with equivalent circuit modeling, Energy Conv. Manag. 87 (2014) discharge cycle for an EV, Appl. Therm. Eng. 135 (2018) 123–132.
367–377. [113] P. Nie, S. Zhang, A. Ran, C. Yang, S. Chen, Z. Li, X. Zhang, W. Deng, T. Liu,
[86] L. Chen, M. Zhang, Y. Ding, S. Wu, Y. Li, G. Liang, H. Li, H. Pan, Estimation the F. Kang, G. Wei, Full-cycle electrochemical-thermal coupling analysis for
internal resistance of lithium-ion-battery using a multi-factor dynamic internal commercial lithium-ion batteries, Appl. Therm. Eng. 184 (2021), 116258.
resistance model with an error compensation strategy, Energy Rep 7 (2021) [114] A.J. Crawford, D. Choi, P.J. Balducci, V.R. Subramanian, V.V. Viswanathan,
3050–3059. Lithium-ion battery physics and statistics-based state of health model, J. Power
[87] N. Damay, C. Forgez, M.P. Bichat, G. Friedrich, Thermal modeling of large Sources 501 (2021), 230032.
prismatic LiFePO4/graphite battery. Coupled thermal and heat generation [115] M.W. Tahir, C. Merten, Multi-scale thermal modeling, experimental validation,
models for characterization and simulation, J. Power Sources 283 (2015) 37–45. and thermal characterization of a high-power lithium-ion cell for automobile
[88] H. Shi, L. Wang, S. Wang, C. Fernandez, X. Xiong, B.E. Dablu, W. Xu, A novel application, Energy Conv. Manag. 258 (2022), 115490.
lumped thermal characteristic modeling strategy for the online adaptive [116] Y. Yang, L. Chen, L. Yang, X. Du, Numerical study of combined air and phase
temperature and parameter co-estimation of vehicle lithium-ion batteries, change cooling for lithium-ion battery during dynamic cycles, Int. J. Therm. Sci.
J. Energy Storage 50 (2022), 104309. 165 (2021), 106968.
[89] C.X. He, Q.L. Yue, M.C. Wu, Q. Chen, T.S. Zhao, A 3D electrochemical-thermal [117] Y. Liu, S. Tang, L. Li, F. Liu, L. Jiang, M. Jia, Y. Ai, C. Yao, H. Gu, Simulation and
coupled model for electrochemical and thermal analysis of pouch-type lithium- parameter identification based on electrochemical-thermal coupling model of
ion batteries, Int. J. Heat Mass Transf. 181 (2021), 121855. power lithium ion-battery, J. Alloy. Compd. 844 (2020), 156003.
[90] Y. Li, F. Qi, H. Guo, Z. Guo, M. Li, W. Wu, Characteristic investigation of an [118] Z. Liu, H. Wang, C. Yang, J. Zhao, Simulation study of lithium-ion battery thermal
electrochemical-thermal coupled model for a LiFePO4/Graphene hybrid cathode management system based on a variable flow velocity method with liquid metal,
lithium-ion battery, Case Stud, Therm. Eng. 13 (2019), 100387. Appl. Therm. Eng. 179 (2020), 115578.
16
[119] F. Chen, W. Zhu, X. Kong, Y. Huang, Y. Wang, Y. Zheng, D. Ren, Study on the [127] C.H. Chen, F.B. Planella, K. O’Regan, D. Gastol, W.D. Widanage, E. Kendrick,
Homogeneity of Large-Size Blade Lithium-Ion Batteries Based on Thermoelectric Development of experimental techniques for parameterization of multi-scale
Coupling Model Simulation, Energies 15-24 (2022) 9556. lithium-ion battery models, J. Electrochem. Soc. 167 (2020), 080534.
[120] Z. Geng, J. Groot, T. Thiringer, A time and cost effective method for entropic [128] J. Li, F. Yang, X. Xiao, M.W. Verbrugge, Y.T. Cheng, Potentiostatic intermittent
coefficient determination of a large commercial battery cell, IEEE Trans. Trans. titration technique (PITT) for spherical particles with finite interfacial kinetics,
Electrif. 6 (2020) 257–266. Electrochim. Acta 75 (2012) 56–61.
[121] S. Shahid, M.A. Chaab, A review of thermal runaway prevention and mitigation [129] M.S. Patil, J.H. Seo, M.Y. Lee, A novel dielectric fluid immersion cooling
strategies for lithium-ion batteries, Energy Conv. Manag X. 16 (2022), 100310. technology for Li-ion battery thermal management, Energy Conv. Manag. 229
[122] X. Feng, D. Ren, X. He, M. Ouyang, Mitigating thermal runaway of lithium-ion (2021), 113715.
batteries, Joule 4 (2020) 743–770. [130] U. Morali, A numerical and statistical implementation of a thermal model for a
[123] G.H. Kim, A. Pesaran, R. Spotnitz, A three-dimensional thermal abuse model for lithium-ion battery, Energy 240 (2022), 122486.
lithium-ion cells, J. Power Sources. 170 (2007) 476–489. [131] O. Mypati, T. Anwaar, D. Mitra, S.K. Pal, P. Srirangam, Characterization and
[124] P.T. Coman, E.C. Darcy, C.T. Veje, R.E. White, Modelling Li-ion cell thermal modelling of Al and Cu busbar during charging and discharging of Li-ion battery
runaway triggered by an internal short circuit device using an efficiency factor for electric vehicles, Appl. Therm. Eng. 218 (2023), 119239.
and arrhenius formulations, J. Electrochem. Soc. 164-4 (2017) A587–A593. [132] T.F. Yang, P.Y. Lin, L.T. Teng, S. Rashidi, W.M. Yan, Numerical and experimental
[125] Y. Li, G. Liu, Z. Li, Numerical modeling of thermal runaway in high-energy study on thermal management of NCM-21700 Li-ion battery, J. Power Sources
lithium-ion battery packs induced by multipoint heating, Case Study, Therm. Eng. 548 (2022), 232068.
38 (2022), 102335. [133] X. Li, J. Zhao, J. Duan, S. Panchal, J. Yuan, R. Fraser, M. Fowler, M. Chen,
[126] G.K. Mertin, D. Wycisk, M. Oldenburger, G. Stoye, A. Fill, K.P. Birke, A.D. Wieck, Simulation of cooling plate effect on a battery module with different channel
Dynamic measurement of the entropy coefficient for battery cells, J. Energy arrangement, J. Energy Storage 49 (2022), 104113.
Storage 51 (2022), 104361. [134] H. Zhang, C. Li, R. Zhang, Y. Lin, H. Fang, Thermal analysis of a 6s4p Lithium-ion
battery pack cooled by cold plates based on a multi-domain modeling framework,
Appl. Therm. Eng. 173 (2020), 115216.
17

1 s2.0 S0017931023008931 Main

Uploaded by

Copyright:

Available Formats

1 s2.0 S0017931023008931 Main

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S0017931023008931 Main

Uploaded by

Copyright:

Available Formats

International Journal of Heat and Mass Transfer 218 (2024) 124748

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer

Review of thermal coupled battery models and parameter identification for

(1) The material properties of the battery are assumed to be inde

Based on these assumptions, the transport phenomena inside the

described by Eq. (3). However, more complex models, such as the ∫t ( )

dUi Ui Ib multiple resistive and capacitive elements to better capture dynamic

Table 1 Table 1 (continued )

(1) Lower the likelihood of hazardous conditions occurring;

Fig. 8. Test process for the identification of entropy coefficient [42].

Fig. 9. HPPC test for the identification of R0 [42].

Ui (t) = Ri (SOC, T)I(t)e− t/Ri (SOC,T)Ci (SOC,T)

The time constant τi=RiCi of each RC network can then be rewritten

Han et al. [41] – 94 0.33C – 3h 5% 25 0 3D CFD

8. Conclusions methodology introduced here could be easily adapted to include

You might also like

1 s2.0 S0017931023008931 Main

Uploaded by

Copyright:

Available Formats

1 s2.0 S0017931023008931 Main

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S0017931023008931 Main

Uploaded by

Copyright:

Available Formats

International Journal of Heat and Mass Transfer 218 (2024) 124748

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer

Review of thermal coupled battery models and parameter identification for

(1) The material properties of the battery are assumed to be inde­

Based on these assumptions, the transport phenomena inside the

described by Eq. (3). However, more complex models, such as the ∫t ( )

dUi Ui Ib multiple resistive and capacitive elements to better capture dynamic

Table 1 Table 1 (continued )

(1) Lower the likelihood of hazardous conditions occurring;

Fig. 8. Test process for the identification of entropy coefficient [42].

Fig. 9. HPPC test for the identification of R0 [42].

Ui (t) = Ri (SOC, T)I(t)e− t/Ri (SOC,T)Ci (SOC,T)

The time constant τi=RiCi of each RC network can then be rewritten

Han et al. [41] – 94 0.33C – 3h 5% 25 0 3D CFD

8. Conclusions methodology introduced here could be easily adapted to include

You might also like

(1) The material properties of the battery are assumed to be inde