The early warning for thermal runaway of lithium-ion batteries based on

internal temperature measurement

Teng Jia, Ying Zhang, Chuyuan Ma, Hang Yu, Sihang Hu
School of Safety Science and Emergency Management, Wuhan University of Technology, Wuhan,

430070, china

HIGHLIGHTS

(1) The internal-external temperature change at different rates of charge and discharge.

(2) The relationship model of internal-external temperature was proposed.

(3) The relationship between internal and external heat transfer coefficients.

(4) A new early warning model for thermal runaway of lithium-ion batteries was

proposed.

Keywords: Lithium-ion batteries; High rate; Charge and discharge; Internal

temperature; Early warning.

ABSTRACT:

It is difficult to monitor the internal temperature of lithium-ion batteries (LIBs),

and surface temperature is generally used as a monitoring parameter. However, there is

a certain difference between the surface temperature and the internal temperature,

especially in the case of high rate charge and discharge. Therefore, it is urgent to

establish a relationship model between internal and external temperatures at different

charge and discharge rates, so as to monitor LIBs in real-time through internal

temperature. In this paper, by embedding thermocouples in the center of the battery’s

negative electrode, the temperature distribution inside and outside the lithium battery

is first discussed. The variation law of temperature and heat generation at different rates
(1-7 C) of charge and discharge is studied, and the relationship model of internal and

external temperature is proposed based on this. By comparing the theoretical and

mathematical models of the internal and external temperatures, the proportional

relationship between the internal and external heat transfer coefficients is determined.

Finally, an early secondary warning model for thermal runaway is proposed based on

the internal and external temperature of the battery, the SEI state and the safe operating

temperature of the battery.

1. Introduction

Since the commercialization of lithium-ion batteries (LIBs) in the early 1990s,

they have been widely used in electric vehicles, energy storage power stations,

aerospace, and other fields due to their advantages of high voltage, high specific energy,

long cycle life and no serious memory effect [1]. During the operation of the battery,

the thermal effect will affect the temperature and electrochemical performance of the

battery, which greatly affects the safety and service life of the battery [2]. At present,

LIBs combustion and explosion accidents occur frequently at home and abroad.

Overheating of the battery during charging/discharging is one of the main causes of

these accidents [3].

LIBs are more likely to overheat during high rate charging and discharging.

Overheating of the lithium battery may cause adverse reactions of the battery materials,

which may lead to further heating of the lithium battery, resulting in thermal runaway

[4]. Electric vehicles require that the battery can withstand relatively high rate

charging/discharging operations. Especially with the emergence of super-fast charging

technology, it has become more urgent to study the thermal characteristics of lithium

battery in the process of high rate charging/discharging. The current temperature

measurement method of LIBs is mainly aimed at the surface temperature [5]. Under

high charge/discharge rates, the temperature difference between the internal and

external of LIBs is obvious [6], and the surface temperature can’t accurately indicate

the safe state of LIBs. Therefore, based on the accurate measurement of the internal and

external temperature, it is very important for effective thermal monitoring and thermal

control to construct the internal and external temperature relationship model of the

lithium battery during charge and discharge.

The research on the internal temperature of lithium battery mainly includes two

parts: experiment and numerical algorithm. In the experimental aspect, the

measurement method of internal temperature is mainly explored, such as embedding

micro thermocouples into the battery to measure the internal temperature [7, 8]. Zhe Li

et al. [8] embedded thermocouples at multiple points in the stacked battery, and the

temporal and spatial changes of internal and external temperatures at different discharge

rates were studied. For in-situ monitoring of internal temperature, fiber Bragg gratings

were implanted into lithium batteries based on their small size and corrosion resistance

[9, 10], so as to realize non-destructive monitoring of internal temperature of batteries.

In numerical algorithm aspect, the real-time prediction of external temperature to

internal temperature was realized by establishing numerical model [11-17]. At present,

the core temperature estimation based on the electro-thermal model is more widely

studied. For example, Lunguo Chen et al. [17] developed a new electro-thermal model
of lithium battery by coupling the lumped parameter double thermal model and

equivalent circuit model. Through the electro-thermal model, an adaptive algorithm of

battery core temperature based on joint Kalman filter was proposed. It can be seen that

the internal temperature mainly focuses on the internal temperature measurement

methods and temperature prediction models, and the model research based on the

internal and external temperature at different rates of charge and discharge conditions

is still relatively rare.

The research on (LIBs) safety monitoring and early warning mainly includes

temperature monitoring, strain monitoring, gas monitoring and other means. Some

studies have implanted fiber grating sensors in lithium battery to monitor the real-time

temperature and strain inside the battery [18-20], but none of them involve monitoring

and early warning of thermal runaway. In terms of gas monitoring, it is mainly aimed

at battery overcharge or thermally induced runaway. Gas monitoring was based on the

gas discharged during thermal runaway [21-23], which is not an early warning of

thermal runaway. Some scholars have also studied the problem that the attached FBG

sensor was susceptible to interference during strain monitoring. For example, Peng et

al. [24] proposed a new high-precision strain monitoring method based on FBG sensors.

In order to improve the accuracy of monitoring and warning, Nascimento et al. [25]

proposed a hybrid sensor network composed of fiber Bragg grating and Fabry-Perot

cavity. As for the battery monitoring means, Micael Nascimento [26] et al. compared

the temperature responses of different parts of the battery by using thermocouple and

fiber grating sensors.

 In summary, in the aspect of battery internal temperature, the current research

mainly focuses on internal and external temperature monitoring and internal

temperature prediction. In terms of battery thermal runaway warning, the current

research mainly focuses on the analysis of battery temperature, deformation, gas and

other parameters before and after thermal runaway, and the early monitoring and

warning of LIBs thermal runaway based on internal temperature has not been found.

As for the thermal runaway monitoring and warning, most studies give the threshold of

warning parameters for the critical conditions of thermal runaway. In fact, when the

battery reaches the critical condition of thermal runaway, the SEI film of the battery

has been destroyed, and the safety performance is obviously decreased. At this time,

the early warning of the battery is already a late warning, and the heat accumulated in

the battery may still cause thermal runaway, especially in the case of high rate charge

and discharge. In order to make the thermal runaway monitoring and warning more

accurate and timely, it is necessary to study the internal and external temperature

changes during battery charging and discharging, and give the thermal runaway early

warning model based on internal temperature. In view of the lack of current research,

this paper carried out research on charge-discharge temperature and heat generation of

LIBs at different rates. According to the internal and external temperature relationship

of different rates, the internal and external temperature relationship model of LIBs is

constructed. Finally, based on the battery performance, SEI state and internal and

external temperature model, a two-level early warning method for thermal runaway of

LIBs is proposed.
2. Experiment

2.1. Charge and discharge experiment of lithium battery

The battery used in this paper is a NCM ternary lithium battery produced by

Samsung. The nominal capacity of the test battery is 2500 mAh. The battery charging

method is constant current and constant voltage charging. The actual capacity of the

battery is obtained by charging and discharging three times. The specific parameters of

the battery are shown in Table 1.

Table 1 Samsung 18650 battery parameters

Charge Discharge
Product Nominal Internal
Weight cut-off cut-off
model capacity resistance
voltage voltage
Samsung
45 g 2500 mAh ≤13.5 mΩ 4.2 V 2.5 V
25R

To test the internal temperature, the thermocouple was inserted into a borehole at

the exact center of the negative electrode with a diameter of 1.5mm. Six K-type

thermocouples were arranged in the experiment. Three thermocouples were placed in

parallel in the borehole, and the battery was re-sealed with aluminum silicate inorganic

binder. Three thermocouples are arranged in parallel on the upper, middle, and lower

parts of the battery surface, respectively. The arrangement of thermocouples inside and

outside the battery is consistent, as shown in Fig.1.

The internal thermocouple is T1in ~ T3in, and the external thermocouple is T1out ~

T3out. In order to reduce the influence of the built-in thermocouple operation on the

battery, the whole operation process is carried out in a dry room. After the thermocouple

is inserted into the borehole, the hole is sealed immediately.

 Thermocouple

T1in
T2in
T3in

T1out
T2out
T3out

(a) (b)

Fig.1. Thermocouple position：

(a) Battery sample; (b) Thermocouple arrangement.

The experimental platform of lithium battery charging and discharging at different

rates is shown in Fig.2. The experimental platform mainly includes charge and

discharge cycle instrument, thermocouple, constant temperature and humidity box, data

acquisition instrument, etc. The data acquisition instrument is used for real-time

acquisition of thermocouple temperature, and the constant temperature and humidity

box is set to 25 ℃. To explore the heat generation characteristics of the battery at high

charge and discharge rates, the charge and discharge rates were set to 1 C, 2 C, 3 C, 4

C, 5 C, 6 C, and 7 C, with a total of 7 working conditions.

Battery testing system

Charge and discharge

Temperature acquisition
Data terminal

Constant temperature and humidity box

Data-acquisition instrument
 Fig.2. Lithium battery charge and discharge experimental platform under

different rate conditions.

The experimental steps of each group are set as follows: (1) The lithium battery

was charged to 100 % SOC under constant current and constant voltage conditions. (2)

The battery was allowed to stand for 1 h to stabilize the voltage and current, and the

surface temperature was restored to 25 ℃. (3) The lithium battery was discharged to 0 %

SOC under constant current conditions. (4) The battery was allowed to stand for 2 h,

and the surface temperature was restored to 25 ℃. The (1)-(4) steps were cycled with

different rate currents until all working conditions were completed. The charge voltage

was set to 4.2 V, and the discharge cut-off voltage was set to 2.5 V. The charge and

discharge currents of 1 C, 2 C, 3 C, 4 C, 5 C, 6 C, and 7 C correspond to 2.5 Ah, 5.0

Ah, 7.5 Ah, 10 Ah, 12.5 Ah, 15 Ah, and 17.5 Ah, respectively.

2.2. Verification of embedded thermocouple method

The specific steps of embedding thermocouple are as follows: (1) The center of

the battery negative electrode was drilled by a drill with a diameter of 1.5 mm. (2) The

voltage was measured to preliminarily check whether a micro-short circuit occurred.

(3) The thermocouples measurement end skin was removed and embedded in parallel

to a specific depth in the hole. (4) The battery was re-sealed using an aluminum silicate

inorganic adhesive. (5) The voltage was measured to detect whether a short circuit

occurred. (6) After the battery was sealed, the electrolyte leakage was observed for

more than 24 hours. (7) Before the experiment, the battery capacity was tested by

charging and discharging instrument to check whether there was obvious capacity loss.
 To evaluate the effect of the embedded thermocouple on the performance of the

primary battery, the capacity and internal resistance of the battery before and after

drilling were compared. The comparison of capacity and internal resistance before and

after drilling is shown in Table 2.

Table 2 Comparison of 1 C charge and discharge capacity and internal resistance

before and after drilling

No internal There are internal Variable

Condition
thermocouples thermocouples quantity

charge capacity
2.48 2.47 0.4%
(Ah)
discharge capacity
2.48 2.46 0.8%
(Ah)
Charge internal
resistance 33 32 3.0%
(mΩ)
Discharge internal
resistance 31 30 3.2%
(mΩ)

It can be seen from Table 2 that the charge and discharge capacity and internal

resistance decreased by 0.4 %, 0.8 %, 3.0 %, and 3.2 % respectively after the battery

was drilled into the thermocouple, which may be caused by the volatilization of the

electrolyte. The experimental results of battery capacity and internal resistance before

and after drilling are shown in supplementary Fig.1 (a) and Fig.1 (b). As can be seen

from supplementary Fig.1, the voltage-capacity curves basically coincided, and the

maximum internal resistance difference was 2.1 mΩ, indicating that the built-in
thermocouple has little impact on the battery performance.

3. Heat generation model

3.1 Calculation of heat generation

The heat generation rate of LIBs generally adopts the Bernardi heat generation

model. As shown in Eq. (1).

dU
q  I (V  U )  IT  Qmi  Q pc (1)
dT

Where q is the heat generation rate of the battery, I is the working current, V

is the working voltage, U is the open circuit voltage, T is the battery temperature,

dU
and is the voltage temperature coefficient. The first term on the right side of Eq.
dT

(1) represents Joule heat, which is irreversible heat. The second term represents

reversible heat, which is mainly caused by entropy change. The third term represents

the heat of chemical reaction, and the fourth term represents the heat of phase change

reaction. In the normal charge and discharge process, the third and fourth terms can be

ignored relative to the reversible heat and irreversible heat. The Eq. (1) usually be

abbreviated to Eq. (2).

dU
q  I (V  U )  IT (2)
dT

To obtain the voltage temperature coefficient, the battery is charged to a pre-set

SOC and placed in a constant temperature box to record the voltage change [27-29].

The voltage temperature coefficient is obtained by calculating the slope of the

temperature-voltage curve.
3.2 Theoretical model of internal and external temperature

The energy balance equation of lithium battery can be expressed as:

Tin
q  Cp  qout (3)
t
Where Cp is the heat capacity, Tin is the internal temperature of the battery,

qout is the heat flow transferred to the outside of the battery. In this paper, in order to

establish a simplified thermal model, we assume that the battery shell temperature and

internal temperature are uniform, and the generated heat is evenly distributed inside the

battery. Because it is a steady heat transfer process, the heat flux through the inside and

outside of the battery should be the same [30]. Fig.3 shows the internal and external

heat transfer of lithium battery.

r R

qin qout

Tin Tout

hin hout
dx
x

Fig.3. The internal and external heat transfer of lithium battery.

According to Fig.3, it can be obtained

dTin 
qin   
dx

qout  hout (Tout  Tamb )  (4)
qin  qout 



Where  is the thermal conductivity of the battery, hout is the external heat
transfer coefficient of the battery, Tout is the surface temperature of the battery, Tamb

is the ambient temperature, qin is the heat flow inside the battery.

The radial temperature of the battery is uniformly distributed, and the temperature
dTin
gradient is constant, which can be obtained
dx
k
 (Tin  Tout )  hout (Tout  Tamb ) (5)
(R  r)

k
Let hin   ，defining hin as the internal heat transfer coefficient of the
(R  r)

battery, we can get

hin T  Tamb
 out (6)
hout Tin  Tout

According to equation (6), if Tout and Tamb are known, then Tin can be

obtained. The internal temperature expression is given

hout h
Tin  Tout (1  )  Tamb out (7)
hin hin

4. Results and discussion

4.1. Temperature distribution law

To explore the temperature distribution of 18650 lithium battery, a new 18650

lithium battery was used to carry out experiments at different rates ( 2 C, 4 C, and 6 C ),

and the internal and external temperature distributions of lithium battery were compared.

Fig. 4 (a), 4 (b), and 4 (c) respectively show the internal and external temperature

changes of 2 C, 4 C, and 6 C during charging.

 1405s 64
704s
40
T1in 4C T1in
2C 60
38 ΔTin(max)=0.27℃ ΔTin(max)=1.01℃
T2in
T2in
56
T3in T3in
36
T1out 52 T1out
Temperature(℃)

Temperature(℃)
34 T2out T2out
48
ΔTout(max)=0.51℃ ΔTout(max)=1.59℃
T3out T3out
32 44

30 40

36
28 Constant current charge Constant voltage charge
32
26 Constant current charge Constant voltage charge
28
24 24
0 500 1000 1500 2000 2500 3000 0 500 1000 1500 2000
Time(s) Time(s)

(a) (b)

501s
78
6C ΔTin(max)=1.47℃ T1in
72
T2in
66 T3in
T1out
60
Temperature(℃)

ΔTout(max)=2.28℃ T2out
54 T3out

48

42

36
Constant current charge Constant voltage charge
30

24
0 200 400 600 800 1000 1200 1400 1600
Time(s)

(c)

Fig.4. Internal and external temperature changes of the upper, middle, and lower

parts of the lithium battery during charging: (a) 2 C; (b) 4 C; (c) 6 C.

As can be seen from Fig. 4, there was a temperature gradient in axial and radial

directions during the charging of the 18650 cylindrical battery, and the temperature

difference in the radial direction was particularly obvious. In Fig. 4 (a), 4 (b), and 4 (c),

the temperature distribution has similar law. According to the temperature analysis of

the three parts of the lithium battery surface, the temperature near the negative electrode

was relatively low, and the temperature of the other two surfaces was similar. The

internal temperature distribution of lithium battery showed the opposite characteristics.

The internal temperature near the negative electrode was relatively high, while the other
two internal temperatures were similar. Overall, the axial surface temperature and

internal temperature gradient are not obvious. The maximum axial temperature

difference of lithium battery at 2 C, 4 C, and 6 C charging rates is shown in Table 3.

Table 3 The maximum axial temperature difference of lithium battery at 2 C, 4 C, and

6 C charging rates

Maximum internal Maximum surface

Maximum
Charging temperature difference temperature difference
temperature time
rate Tin (max) Tout (max)
(s)
(℃) (℃)

2C 1450 0.27 0.51

4C 704 1.01 1.29
6C 501 1.47 2.28

As shown in Table 3, with the increase of the rate, the temperature difference

between surface temperature and internal temperature of lithium battery increased, but

the overall temperature difference was relatively small. The axial temperature gradient

of lithium battery is not obvious under different rate conditions. The average

temperature of the three parts can be selected for internal and external temperature

analysis.

4.2. Temperature variation characteristics of charge and discharge at different

rates

To make the experiment closer to reality, the same kind of battery with certain

aging was selected for high rate charge and discharge experiments. The selected battery

capacity was 2400 mAh.

4.2.1. Internal and external temperature analysis of charge and discharge

The temperature variation characteristics of lithium battery charge and discharge

are shown in Fig.5. Fig. 5 (a), 5 (b), 5 (c), and 5 (d) are the internal and external

temperature variation curves of lithium battery charge and discharge. Figs.5 (e) and 5

(f) are the internal and external temperature comparison curves of lithium battery

charge and discharge.

55 70
60
65 70
Tin-1C
Maximum external temperature 53.17 Tout-1C Maximum internal temperature
66.33

50 50 49.52 60
60.89
Tin-2C

Maximum temperature(℃)
Tout-2C
Maximum temperature(℃)

44.95 54.34
40.76
60 50 47.94
40
35.97 Tout-3C 40.94
Tin-3C
40
45 30.92
Tout-4C 55 Tin-4C
Temperature(℃)
Temperature(℃)

30 27.08 33.28
30 28.55

20 Tout-5C Tin-5C
50 20
Tout-6C Tin-6C
40 10 10
Tout-7C 45 Tin-7C
0 0
1C 2C 3C 4C 5C 6C 7C 1C 2C 3C 4C 5C 6C 7C
Rate
35 Rate
Charge 40 Charge
35
30
30

25 25
0 1000 2000 3000 4000 5000 0 1000 2000 3000 4000 5000
Time(s) Time(s)

(a) (b)
65 85
70
Tout-1C 80 80 78.69
Tin-1C
Maximum external temperature 61.52 Maximum internal temperature
60 60 70.82
55.78
Tout-2C 75
70
Tin-2C
Maximum temperature(℃)
Maximum temperature(℃)

49.38 61.27
50 60
43.79 Tout-3C 70
52.83 Tin-3C
55 40 38.25 50
44.37
33.28 Tout-4C 40 Tin-4C
30 28.55 65 36.78
Temperature(℃)
Temperature(℃)

50 Tout-5C 30
30.42
Tin-5C
20
Tout-6C 60 20
Tin-6C
10 10
45 Tout-7C 55 Tin-7C
0 0
1C 2C 3C 4C 5C 6C 7C 1C 2C 3C 4C 5C 6C 7C
Rate
50 Rate
40 Discharge Discharge
45
35 40
35
30
30
25 25
0 500 1000 1500 2000 2500 3000 3500 4000 0 500 1000 1500 2000 2500 3000 3500 4000
Time(s) Time(s)

(c) (d)
 70 85
80 Tin-1C
65 70 66.33
Tin-1C 80 78.69
Charge-maximum external temperature Discharge-maximum external temperature
60
Charge-maximum internal temperature Tin-3C 75 70 Discharge-maximum internal temperature Tin-3C

Maximum temperature(℃)

Maximum temperature(℃)
60 54.34 53.17
Tin-5C
61.27 61.52
Tin-5C
50 70 60
44.95
40.94 Tin-7C 50 49.38 Tin-7C
55 40 65 44.37

Temperature(℃)
Temperature(℃)

35.97
Tout-1C 40 38.25
Tout-1C
30 27.08 28.55
60 30 28.55 30.42
50 Tout-3C Tout-3C
20
Tout-5C 55 20
Tout-5C
10
45 10
Tout-7C
Tout-7C 50 0
0
1C 3C 5C 7C 1C 3C 5C 7C
40 Rate
45 Rate
Charge Discharge
40
35
35
30
30
25 25
0 1000 2000 3000 4000 5000 0 500 1000 1500 2000 2500 3000 3500 4000
Time(s) Time(s)

(e) (f)

Fig.5. The internal and external temperature changes of lithium battery charge and

discharge at different rates: (a) 1 C-7 C charge external temperature; (b) 1 C-7 C

charge internal temperature; (c) 1 C-7 C discharge external temperature; (d) 1 C-7 C

discharge internal temperature; (e) Comparison of internal and external temperature at

1 C, 3 C, 5 C, and 7 C charge; (f) Comparison of internal and external temperature at

1 C, 3 C, 5 C, and 7 C discharge.

It can be seen from Fig. 5 (a), 5 (b), 5 (c), and 5 (d) that the variation trend of

charge and discharge temperature at different rates during charging and discharging

was generally consistent. With the increase in charge and discharge rate, the

temperature also increased. During the charge of lithium battery, the internal and

external temperatures increased first and then decreased. The temperature in the

constant current charging stage had been rising. The temperature in the constant voltage

charging stage dropped significantly and finally approached the normal temperature.

During the discharge of lithium battery, the internal and external temperatures had been

rising. At the end of constant current discharge, the temperature reached the maximum.

Through Figs.5 (e) and 5 (f), it is found that the change of internal and external
temperature during charge and discharge was basically synchronous. With the increase

of charge and discharge rate, the temperature difference between inside and outside was

increasing. The temperature change wasn’t obvious at 1 C charge and discharge rate,

but the charge and discharge temperature had increased significantly at the other rate.

At the same time, the internal and external temperature changes in the discharge stage

were greater than those in the charge stage. Supplementary Fig.2 compares the internal

and external temperature changes at different rates.

In supplementary Fig.2 (a), during the 1 C-7 C charging stage, the maximum

internal and external temperatures reached 66.33 ℃ and 53.17 ℃ respectively. At 1

C-7 C discharge stage, the maximum internal and external temperatures reached 78.69 ℃

and 61.52 ℃ respectively. In supplementary Fig.2 (b), during the 1C-7C charging stage,

the maximum internal temperature rise increased from 2.44 °C to 39.54 °C, and the

maximum external temperature rise increased from 1.66 °C to 27.43 °C, which

increased by 37.1 °C and 25.77 °C respectively. At 1 C-7 C discharge stage, the

maximum internal temperature rise increased from 4.88 °C to 52.57 °C, and the

maximum external temperature rise increased from 3.05 °C to 36.15 °C, which

increased by 47.69 °C and 33.1 °C, respectively. In supplementary Fig.2 (c), during the

1 C-7 C charging stage, the maximum temperature difference between internal and

external increased from 1.47 °C to 13.16 °C. At 1 C-7 C discharge stage, the maximum

temperature difference between internal and external increased from 1.87 °C to

17.17 °C. The maximum temperature difference in the discharge stage was always

greater than that in the charge stage, and the temperature difference between the two
reached 4.01 ℃ at 7 C.

4.2.2. Heat generation analysis of charge and discharge

The internal resistance of lithium battery was tested by HPPC, and the voltage

temperature coefficient of lithium battery was tested by step heating method. The

internal resistance and voltage temperature coefficient of the battery under different

SOC conditions are shown in supplementary Fig.3.

During high rate charge and discharge, irreversible heat plays a dominant role.

Fig.6 shows the comparison of irreversible heat generation rate under different rate

conditions. Among them, Fig.6 (a) is the comparison of irreversible heat generation rate

during charge, and Fig.6 (b) is the comparison of irreversible heat generation rate during

discharge.

25 25
3102s 0.55
0.5

1C 1C 0.50 1C 1C
Irreversible heat generation rate(W)
Irreversible heat generation rate(W)

Irreversible heat generation rate(W)

Irreversible heat generation rate(W)

20
0.4
2C 20 0.45 2C
3C 0.40 3C
0.3
4C 0.35 4C
Constant current charge Constant voltage charge

15 0.2 5C 15 0.30 5C
6C 0.25 6C
0.1
7C 0.20 7C
0.15
10 0.0 10 0 500 1000 1500 2000 2500 3000 3500 4000
t(s)
0 1000 2000 3000 4000 5000
t(s)

5 Charge 5 Discharge

0 0

0 1000 2000 3000 4000 5000 0 1000 2000 3000 4000 5000
t(s) t(s)

Fig.6. Comparison of irreversible heat generation rate of charge and discharge under

different rate conditions: (a) Charge; (b) Discharge.

It can be seen from Fig. 6 that the irreversible heat change trend of lithium battery

under different charge and discharge rates was basically the same. In Fig.6 (a), the

irreversible heat during charging showed a trend of decreasing first, then increasing,
and then decreasing rapidly. The trend of first down and then up was mainly reflected

in the constant current charging stage, and the irreversible heat was dominated by the

internal resistance change. The rapid decline was reflected in the constant current stage,

and the irreversible heat was mainly dominated by the current change. In Fig.6 (b), the

irreversible heat during discharging generally showed a trend of decreasing first and

then increasing, and the irreversible heat was mainly dominated by internal resistance.

By comparing the heat generation rate of charge and discharge at different rates, it can

be found that irreversible heat increases with the increase of charge and discharge rate.

Supplementary Fig.4 is the comparison of charge and discharge heat generation. Where

Q1 is the heat generation of charging, Q2 is the heat generation of discharging,

Q  Q2  Q1 is the heat generation difference between charging and discharging.

In supplementary Fig.4, as the charge and discharge rate increases, the heat

generation increased, and the difference between charge and discharge heat generation

also increased. The heat generation of discharge was always greater than that of

charging, and there were differences in heat generation at different rates. The heat

generation of 1 C charge and discharge was not much different, and it could be

considered that the heat generation of 1 C charge and discharge was approximately

equal. When charging and discharging above 2 C rate, the heat generation by

discharging was much larger than that by charging. When charging and discharging at

7 C, the heat generation difference reached 1068 J. Comparing the temperature and heat

generation of charge and discharge at different rates, it could be found that the

temperature and heat generation change law were consistent.

4.3. Internal and external temperature relationship

The internal and external temperature data at different charge and discharge rates

are fitted, as shown in Fig.7. In Fig.7, there was a good linear relationship between the

internal temperature and the external temperature.

80

70

60
Tin（℃）

50
Tin  11.97415  1.4769Tout
40 R 2  0.99974

30

20 Relationship between internal and external temperature

20 25 30 35 40 45 50 55 60 65
Tout（℃）

Fig.7. The relationship between internal and external temperature at different charge

and discharge rates.

The relationship between internal and external temperature is fitted as:

Tin  11.974  1.477Tout (8)

When Tamb =25℃ , By transforming Equation (8), we can obtain

Tin  (1+0.477)Tout  0.47896Tamb (9)

It can be found that Equation (7) and Equation (9) have good consistency.

Combining Eqs. (7) and (9), the relationship between internal and external heat transfer

coefficients can be obtained

hout
 0.48 (10)
hin

The significance of equation (10) is that the ratio of internal and external heat

transfer coefficient can be regarded as a certain value for the heat generation behavior
of lithium battery during charging and discharging.

4.4. The early warning model of thermal runaway

The SEI reaction temperature inside the battery is low, and its decomposition is

generally considered to be the initial exothermic reaction of the battery thermal

runaway. The decomposition of SEI will produce a lot of heat and even trigger the next

chain reaction, which is an important factor causing the thermal runaway of the battery.

Therefore, the state of SEI is an important reference in the early warning of lithium

battery thermal runaway.

Related research shows that when the surface temperature reached 80 °C, the

calorimeter detected obvious exothermic phenomenon, which means that SEI begins to

decompose and release heat [31, 32]. Literature [33] pointed out that typical SEI

exothermic decomposition reaction started at 60-120℃ through the temperature

diagram of lithium battery thermal runaway process. By DSC (Differential Scanning

Calorimetry) test, Richard et al. found that SEI had an obvious exothermic peak at about

100 °C [34, 35], indicating that SEI had begun to decompose and release heat at about

100 °C. The SEI decomposition temperature can be used as an early warning

temperature. When the internal temperature is 100 ℃, the corresponding external

temperature of the battery is 76 ℃. Based on the consideration of early warning, 80°

C is finally selected as the early warning temperature.

The formation and growth of SEI have a great dependence on temperature. Higher

temperature accelerates the growth of SEI, affects the stability of SEI, and leads to the

increase of battery internal resistance and capacity loss. In the high temperature
environment, even intermittently, the battery will cause irreversible chemical

conversion in the passivation layer, and its organic components tend to be converted

into inorganic materials[36]. These processes help deplete the Li+ pool in the battery,

and they are major contributors to the self-discharge mechanism [37, 38]. Lin Liu et al.

[39] found that the thickness of SEI film increased significantly at high temperature and

high rate. When the ambient temperature was 45 ℃, the thickness of SEI film in the

first 8 cycles increased by more than 15 % compared with room temperature. Literature

[40] also proposed that SEI cycle formation above 40℃ is detrimental to the long-term

cycle performance of LIBs. Therefore, the ambient temperature affecting SEI formation

and growth can be used as an early warning temperature. For 18650 lithium battery, the

working temperature is generally maintained at -20 ℃ ～ 50 ℃, and the optimum

working temperature is 20 ℃ ～ 40 ℃. Considering the growth of SEI and the safe

working temperature of the battery, 50 ℃ is selected as the initial warning temperature.

In this paper, combined with the battery safety working temperature, SEI state and

temperature model, a hierarchical early warning method for thermal runaway of lithium

battery charging and discharging is proposed, as shown in table 3. Fig.8 is the two-level

warning signal of lithium battery.

Table 3 The early temperature warning signal for thermal runaway of LIBs

Internal temperature External temperature

Early warning stage
(℃) (℃)

First-level 62 50
Second-level 106 80
 120
External temperature
110
Internal temperature
100 106℃

90
Safety area Early waring area Dange area

Temperature(℃)
80
Second signal
70 TW2=80℃
62℃
60

50
First signal
40 TW1=50℃

30

20
-100 0 100 200 300 400 500 600 700
Time(s)

Fig.8. The two-level warning signals for thermal runaway of LIBs

The two-level early warning model of thermal runaway shown in Fig. 8 is

described in detail as follows.

The first-level warning: After the temperature of the lithium battery reaches the

first-level warning value (Tw1=50℃), the continuous charging and discharging affects

the cycle performance of the battery. It is recommended to take cooling measures and

use the thermal management system to dissipate heat in time.

The second-level warning: After the temperature of lithium battery reaches the

second-level warning value (Tw2=80℃), the continuous charging and discharging may

cause thermal runaway. Stop charging and discharging, the battery is not recommended.

5. Conclusions

In this paper, the thermocouples were embedded in the negative electrode of the

cylindrical 18650 lithium battery. The charge and discharge temperature and heat

generation characteristics of lithium battery were studied by charging and discharging

experiments at different rates. On this basis, the relationship model of internal and
external temperature was obtained. Combined with battery performance, SEI state and

internal and external temperature relationship model, a two-level early warning method

for lithium battery thermal runaway was proposed.

The main conclusions are as follows:

(1) The charging experiment at different rates (2 C, 4 C, and 6 C) were carried out,

and the axial temperature gradient of lithium battery was found to be small. The

maximum temperature differences inside the axial direction were 0.27 °C at 2 C,

1.01 °C at 4 C, and 1.47 °C at 6 C. The maximum temperature differences outside the

axial direction were 0.51 °C at 2 C, 1.29 °C at 4 C, and 2.28 °C at 6 C. The temperature

inside and outside the experiment was determined to take the average temperature of

different parts of the battery.

(2) The charging and discharging experiment at different rates (1 C-7 C) were

carried out, the discharge temperature of the lithium battery was found to be higher than

the charging temperature. During 7 C charging and discharging, the maximum

temperatures inside and outside the charging were 66.33 °C and 53.17 °C, respectively.

The maximum temperatures inside and outside the discharging were 78.69 °C and 61.52

°C, respectively. The internal temperature difference of charging and discharging

reached 12.36 °C. During the charge and discharge of 1 C-7 C, the maximum

temperature difference in the charging stage increased from 1.47 °C to 13.16 °C, and

the maximum temperature difference in the discharge stage increased from 1.87 °C to

17.17 °C.

(3) With the increase of charge and discharge rate, the heat generation increased,
and the discharge heat generation was always greater than the charge heat generation.

The heat generation of 1 C charge and discharge was roughly flat, and the heat

generation difference becomes larger when the charge and discharge rate was above 2

C. When charging and discharging at 7 C, the heat generation difference reached 1068

J.

(4) The internal and external temperature of 18650 cylindrical lithium battery had

a good linear relationship. According to the internal and external temperature

relationship model, the ratio of internal and external heat transfer coefficients was

determined.

(5) A two-level early warning model of thermal runaway based on internal

temperature measurement was proposed. The first-level warning temperature was 50 ℃

for the battery, and the second-level warning temperature was 80℃ for the battery.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or
personal relationships that could have appeared to influence the work reported in this
paper.

CRediT authorship contribution statement

Conceptualization, investigation, and writing (original draft preparation), Teng Jia;

funding acquisition, project administration, resources, writing (review and editing), and
supervision, Ying Zhang; investigation and resources, Chuyuan Ma; writing (review
and editing), and supervision, Hang Yu, Sihang Hu.

Acknowledgment

This research was financially supported by the National Natural Science

Foundation of China (Nos. 52074202, 52104211), the Fundamental Research Funds for
the Central Universities (Nos: 2022IVA088 and 2022IVA126), and the Key R&D
Program of Hubei Province (2021BCA151).

References

1. Wei, Z., et al., Future smart battery and management: Advanced sensing from external

to embedded multi-dimensional measurement. Journal of Power Sources, 2021. 489.

2. Zhu, S., et al., Data-driven analysis on thermal effects and temperature changes of

lithium-ion battery. Journal of Power Sources, 2021. 482.

3. Abada, S., et al., Safety focused modeling of lithium-ion batteries: A review. Journal of

Power Sources, 2016. 306: p. 178-192.

4. Dong, T., P. Peng, and F. Jiang, Numerical modeling and analysis of the thermal behavior

of NCM lithium-ion batteries subjected to very high C-rate discharge/charge operations.

International Journal of Heat and Mass Transfer, 2018. 117: p. 261-272.

5. Zhang, C., K. Li, and J. Deng, Real-time estimation of battery internal temperature based

on a simplified thermoelectric model. Journal of Power Sources, 2016. 302: p. 146-

154.

6. Richardson, R.R., P.T. Ireland, and D.A. Howey, Battery internal temperature estimation

by combined impedance and surface temperature measurement. Journal of Power

Sources, 2014. 265: p. 254-261.

7. Yang, X., et al., Experimental study on temperature difference between the interior and

surface of Li[Ni1/3Co1/3Mn1/3]O2 prismatic lithium-ion batteries at natural

convection and adiabatic condition. Applied Thermal Engineering, 2021. 190.

8. Li, Z., et al., Examining temporal and spatial variations of internal temperature in large-
 format laminated battery with embedded thermocouples. Journal of Power Sources,

2013. 241: p. 536-553.

9. Yang, G., et al., Real-time temperature measurement with fiber Bragg sensors in lithium

batteries for safety usage. Measurement, 2013. 46(9): p. 3166-3172.

10. Novais, S., et al., Internal and External Temperature Monitoring of a Li-Ion Battery with

Fiber Bragg Grating Sensors. Sensors, 2016. 16(9): p. 9.

11. Sun, L., W. Sun, and F. You, Core temperature modelling and monitoring of lithium-ion

battery in the presence of sensor bias. Applied Energy, 2020. 271.

12. Wang, Q.-K., et al., Decoupling parameter estimation strategy based electrochemical-

thermal coupled modeling method for large format lithium-ion batteries with internal

temperature experimental validation. Chemical Engineering Journal, 2021. 424.

13. Kleiner, J., et al., Real-time core temperature prediction of prismatic automotive lithium-

ion battery cells based on artificial neural networks. Journal of Energy Storage, 2021.

39.

14. Jin, C., et al., Model and experiments to investigate thermal runaway characterization

of lithium-ion batteries induced by external heating method. Journal of Power Sources,

2021. 504.

15. Zhu, J.G., et al., A new lithium-ion battery internal temperature on-line estimate method

based on electrochemical impedance spectroscopy measurement. Journal of Power

Sources, 2015. 274: p. 990-1004.

16. Qin, P., et al., A novel algorithm for heat generation and core temperature based on

single-temperature in-situ measurement of lithium ion cells. Journal of Power Sources,

 2022. 542.

17. Chen, L., et al., Core temperature estimation based on electro‐thermal model of lithium‐

ion batteries. International Journal of Energy Research, 2020. 44(7): p. 5320-5333.

18. Nascimento, M., et al. Lithium batteries temperature and strain fiber monitoring. in 24th

International Conference on Optical Fibre Sensors. 2015. SPIE.

19. Fortier, A., et al., Preliminary study on integration of fiber optic Bragg grating sensors in

li-ion batteries and in situ strain and temperature monitoring of battery cells. Energies,

2017. 10(7): p. 838.

20. Yu, Y., et al., Distributed internal thermal monitoring of lithium ion batteries with fibre

sensors. Journal of Energy Storage, 2022. 50: p. 104291.

21. Fernandes, Y., A. Bry, and S. De Persis, Identification and quantification of gases

emitted during abuse tests by overcharge of a commercial Li-ion battery. Journal of

power sources, 2018. 389: p. 106-119.

22. Jin, Y., et al., Detection of micro-scale Li dendrite via H2 gas capture for early safety

warning. Joule, 2020. 4(8): p. 1714-1729.

23. Huang, W., et al., Early warning of battery failure based on venting signal. Journal of

Energy Storage, 2023. 59: p. 106536.

24. Peng, J., et al., High precision strain monitoring for lithium ion batteries based on fiber

Bragg grating sensors. Journal of Power Sources, 2019. 433: p. 226692.

25. Nascimento, M., et al., Internal strain and temperature discrimination with optical fiber

hybrid sensors in Li-ion batteries. Journal of Power Sources, 2019. 410: p. 1-9.

26. Nascimento, M., M.S. Ferreira, and J.L. Pinto, Real time thermal monitoring of lithium
 batteries with fiber sensors and thermocouples: A comparative study. Measurement,

2017. 111: p. 260-263.

27. Liu, J., et al., Heat generation and thermal runaway of lithium-ion battery induced by

slight overcharging cycling. Journal of Power Sources, 2022. 526.

28. Schuster, E., et al., Thermal behavior and electrochemical heat generation in

a commercial 40 Ah lithium ion pouch cell. Journal of Power Sources, 2015. 286: p.

580-589.

29. Onda, K., et al., Thermal behavior of small lithium-ion battery during rapid charge and

discharge cycles. Journal of Power Sources, 2006. 158(1): p. 535-542.

30. Forgez, C., et al., Thermal modeling of a cylindrical LiFePO4/graphite lithium-ion battery.

Journal of Power Sources, 2010. 195(9): p. 2961-2968.

31. D. D. MacNeil, a.D.L., b,* and J. R. Dahna,b,**,z, Comparison of the Reactivity of

Various Carbon Electrode Materials with Electrolyte at Elevated Temperature. Journal of

The Electrochemical Society, 1999. 146(10): p. 3596-3602.

32. Leising, R.A., et al., Abuse testing of lithium-ion batteries - Characterization of the

overcharge reaction of LiCoO2/graphite cells. Journal of the Electrochemical Society,

2001. 148(8): p. A838-A844.

33. Son, K., et al., Thermal and chemical characterization of the solid-electrolyte interphase

in Li-ion batteries using a novel separator sampling method. Journal of Power Sources,

2019. 440: p. 227083.

34. Richard, M.N. and J.R. Dahn, Accelerating Rate Calorimetry Study on the Thermal

Stability of Lithium Intercalated Graphite in Electrolyte. I. Experimental. Journal of The

 Electrochemical Society, 1999. 146(6): p. 2068.

35. Ryou, M.-H., et al., Effects of lithium salts on thermal stabilities of lithium alkyl

carbonates in SEI layer. Electrochimica Acta, 2012. 83: p. 259-263.

36. Verma, P., P. Maire, and P. Novák, A review of the features and analyses of the solid

electrolyte interphase in Li-ion batteries. Electrochimica Acta, 2010. 55(22): p. 6332-

6341.

37. Levi, M., C. Wang, and D. Aurbach, Self-discharge of graphite electrodes at elevated

temperatures studied by CV and electrochemical impedance spectroscopy. Journal of

The Electrochemical Society, 2004. 151(5): p. A781.

38. Utsunomiya, T., et al., Self-discharge behavior and its temperature dependence of

carbon electrodes in lithium-ion batteries. Journal of power sources, 2011. 196(20): p.

8598-8603.

39. Liu, L., et al., A thermal-electrochemical model that gives spatial-dependent growth of

solid electrolyte interphase in a Li-ion battery. Journal of power sources, 2014. 268: p.

482-490.

40. Rodrigues, M.-T.F., et al., High-temperature solid electrolyte interphases (SEI) in

graphite electrodes. Journal of power sources, 2018. 381: p. 107-115.