Articles
Influence of Aerogel Felt with Different Thickness on Thermal Runaway Propagation of 18650 Lithium-ion Battery
2022 Volume 90 Issue 8 Pages 087003
Details
2022 Volume 90 Issue 8 Pages 087003
With the improvement of lithium-ion batteries in civil aviation transportation, the thermal safety of lithium-ion batteries can not be ignored. Especially in a battery pack, the thermal runaway of batteries can spread from cell to cell, resulting in catastrophic hazards. This work focuses on the experimental setup and analysis of the experimental parameters of lithium-ion batteries with different thicknesses of aerogel felt to study the blocking effect of barrier materials on thermal runaway propagation of lithium-ion batteries in civil aviation transport. The aerogel felt was selected as the barrier material and a series of experiments were carried out with different thicknesses of 1 mm, 3 mm, and 6 mm. The results demonstrated that the increase of aerogel felt thickness exhibited excellent performance in delaying lithium-ion battery thermal runaway. Additionally, a simplified thermal model of thermal runaway propagation was proposed to explain the thermal runaway propagation in the battery to adjacent batteries. These results provide valuable suggestions and enlightenment for the aviation safety transportation of lithium-ion batteries.
The lithium-ion battery has attracted widespread attention due to its light mass, high energy density and no memory effect and so on.1–3 With the increasing demand for lithium-ion batteries in the electric vehicle industry, the safety of lithium-ion batteries during civil aviation transportation is particularly important. Although there are many regulations related to the safety of lithium-ion batteries during civil aviation transportation, the accidents of lithium-ion batteries occur frequently due to their danger and the particularity of the civil aviation transportation environment. Therefore, it is necessary to have a thorough understanding of the fire and explosion hazards of lithium-ion batteries.
In lithium-ion battery arrays, thermal runaway may propagate from a failing battery to neighboring batteries and grow into a large-scale fire,4 fire and explosion resulting from thermal runaway and the toxic gas emissions pose huge potential hazards.5 In the process of thermal runaway, the intensity of thermal runaway of the lithium-ion battery varies with the state of charges (SOCs). Battery with higher electrical capacity demonstrated a higher tendency to experience thermal runaway with shorter induction time and resulted in a more energetic response, as indicated by higher maximum temperature rise.6 During the whole process of temperature rise, the battery undergoes the main chain reactions of SEI decomposition, the reaction between the anode and the electrolyte, melting of PE, the decomposition of NCM cathode, and the decomposition of electrolyte, etc.7,8 Complex chemical reactions in lithium-ion batteries lead to various thermal runaway behaviors, which in turn lead to different fire behaviors.
Research on the thermal runaway propagation between lithium-ion batteries has been conducted in detail. Thermal runaway propagation between batteries is affected by many factors, such as battery spacing, abuse conditions, and the arrangement of batteries, mainly due to SOC,9 differences in internal materials, pack structure, and environmental factors. High SOC is attributed to a decrease in the initial temperature of the lithium-ion battery.10 Lithium-ion batteries with different types of internal materials, including an anode, cathode, separator, and full cell (mixing of the three materials including additional electrolyte), have been systematically studied in the accelerating rate calorimetry, differential scanning calorimetry, and C80 micro-calorimeter.11–13 The results showed that the thermal stability of internal material for tested cells follows the below order: anode < separator < cathode.14 Wang calculated the apparent activation energies of LixC6 and LixC6-electrolytes under different SOCs and found that the activation energies decreased with the increase of intercalated lithium.15 The relationship between thermal runaway propagation and external heat source was revealed by Gao.16 Jia investigates the characteristics of thermal runaway propagation for the LiFePO4 and LiNi0.5Co0.2Mn0.3O2 modules at 95, 70, and 35 kPa. The results indicated that TR behaviors become weaker and the average maximum temperature of modules decreases 20–50 °C as the pressure decreases.17 Besides, Tang found that the cells directly connected to thermal runaway cells are usually thermally conductive, while radiation heat transfer is not. And increasing spacing can prevent thermal runaway propagation by the three heat transfer modes.18 We studied the characteristics of lithium ion battery with thermal runaway in different pressure. The results showed that, compared with 96 kpa, there is a greater impact on the ignition and explosion of lithium ion battery under 61 kPa, and the thermal runaway temperature of battery is higher, the explosion injection pressure is lower.19
With the increasing energy density of lithium-ion batteries, the problem of battery safety transportation is becoming more and more serious. Once the cell thermal runaway occurred, cell-to-cell thermal runaway propagation can result in catastrophic hazards.20 Therefore the prevention of thermal runaway propagation must be considered in the design of battery and battery pack transport packaging. According to the previous study, Yuan found that the protective effect of different interstitial substances on adjacent cells was significantly different through the combination of simulation and experiment.21 Chen used an epoxy resin board (ERB) to block the thermal runaway propagation of lithium-ion batteries. The results showed that ERB can reduce the maximum temperature of the thermal runaway battery module, reduce the intensity of the thermal runaway battery, and avoid the generation of jet flame.22 Weng found that the existence of aerogel felt could significantly inhibit the development of thermal runaway propagation of battery modules through experiments.23 In conclusion, although previous studies on the thermal runaway propagation of battery packs and the spacer materials of battery packs have been carried out, there has not been a comprehensive study on the transportation status of lithium-ion batteries in the actual transportation process.
Given that the corrugated paper diaphragm used in the transportation package of the lithium-ion battery has a poor blocking effect on the thermal runaway propagation of lithium-ion batteries, a safer barrier material is essential to increase the safe transportation of lithium-ion batteries. In this work, aerogel felt was selected as a barrier material during the civil aviation transportation for 18650 lithium-ion batteries. During the experiment, the battery pack is placed in special cartons made of corrugated paper along with 1 mm, 3 mm and 6 mm aerogel felt to study the aerogel felt’s blocking effect on thermal runaway propagation. In addition, we choose lithium-ion batteries with 30 % SOC that meet the transport standards of lithium-ion batteries and lithium-ion batteries with 100 % SOC that have severe thermal runaway behavior. The experimental parameters of the two kinds of charged lithium-ion batteries under different barrier thicknesses were compared to study the barrier effect of aerogel felt. The results will provide theoretical support for the prevention and extinguishing of thermal runaway accidents and new insights for the design of lithium-ion battery packaging during civil aviation transportation.
The samples used in this study are the type of 18650, with a nominal capacity of 2600 mAh, and the anode and cathode material of graphite/LiNixCoyMn1−x−yO2. The electrolyte is the solution of LiPF6 and the mixture of ethylene carbonate (EC), methyl carbonate (EMC), and dimethyl carbonate (DMC). The separator of the battery is one layer of polyethylene (PE). And the nominal voltage of the battery is 3.7 V, the working voltage is 2.75–4.2 V. Before the test, a lithium-ion battery charge and discharge system (CT2016D, produced in Shenzhen Jingweiye Electronics Co., LTD) was used to charge the battery to 3.0 V at a constant current of 0.5 C, stand for 1 h, then charge it to 4.2 V at a constant current of 0.5 C, and charge it to 0.05 C at a constant voltage, finally adjust the SOC of the battery to 100 %. The charged battery is left to rest for 12 hours at room temperature to ensure chemical stability.
The barrier materials used in this study are the aerogel felt (produced in Langfang Keye Fire Insulation Co., LTD) with glass fiber as a carrier. The thermal conductivity of aerogel felt is low, about 0.017–0.023 W m−1 k−1, the effect of heat insulation is 2–5 times that of traditional heat insulation materials. The aerogel felt has high compressive strength, which did not decompose and deteriorate and Its combustion performance is class A. In addition, it has excellent thermal insulation, and the most important thing is that it is environmentally friendly. In this study, aerogel felt was cut into 1 mm, 3 mm, and 6 mm pieces of the same height and width equal to the lithium-ion battery and placed between the batteries.
2.2 Experimental platformExperiments were conducted in a self-designed combustion and explosion cabin, which is shown in Fig. 1. The size of the cabin is a 0.5 m × 0.5 m × 0.5 m cube with a transparent door and a gas collecting hood of which the slope height is 0.15 m. The external heat source is a 100 W Hongwei heating rod to trigger the thermal runaway propagation, and it stops heating after the thermal runaway of battery A. The data acquisition system includes a paperless recorder, a high-precision electronic scale, a gas analyzer, and a camera. The paperless recorder (RX6048C, produced in Hangzhou Meihong Automation Technology Co., LTD) collects the temperature at a frequency of 0.1 s. The electronic scale with an accuracy of 0.1 g is used to record the mass loss. The gas analyzer (OPTIMA7, produced by a Famous German company) with a pump is used to monitor the variation of gas concentration (O2, CO, and CO2) with an accuracy of 1 %. The camera is used to record the typical experimental phenomena in the whole thermal runaway process. The lithium-ion battery layout is given in Fig. 2. A 2 × 2 battery pack was fixed in the bracket. The thermocouples (WRN-100, produced in Shanghai Kejie Instrument Factory) with a measuring range of 0–1000 °C were fixed at the center of each battery surface to monitor the temperature variation of individual batteries in the process of thermal runaway propagation. Each experiment was repeated at least three times, and effective tests were selected to reduce the experimental error.
Experimental Platform: Experiments were conducted in a self-designed combustion and explosion cabin. The yellow carton in the middle of the platform is made of corrugated paper. The external heat source is a heating rod to triggers the thermal runaway propagation, and it stops heating after the thermal runaway of battery A. The paperless recorder is used to record the temperature. The electronic scale is used to record the mass loss. The gas analyzer with a pump is used to monitor the variation of gas concentration. The camera is used to record the typical experimental phenomena in the whole thermal runaway process.
Lithium-ion batteries layout: the figure above shows the arrangement of lithium batteries and the placement of aerogel felts, thermocouples, and heating rods. The figure on the left is a top view of a lithium-ion battery with aerogel felt during the experiments. The figure on the right is a side view of a lithium-ion battery with aerogel felt during the experiments, and the yellow carton is made of corrugated paper.
In this paper, the role of battery A is equal to the heating rod. And to analyze the explosion of batteries B, C, and D, a simplified heat transfer model is established based on Eq. 1.24,25
| \begin{equation} \rho C_{\text{p}}\frac{\partial T}{\partial t} = \vec{\nabla} (\lambda \cdot \vec{\nabla} T) + Q \end{equation} | (1) |
The temperature of the lithium-ion battery rises mainly because of heat generated by the chemical reaction of electrode materials, Joule heat of short circuit in the battery, and combustion heat caused by battery combustion. The heat of formation of each side reaction is regarded as the internal heat source with the rising of temperature. The order of exothermic reactions in the battery is the SEI film decomposition reaction on the cathode surface, an exothermic reaction between the cathode and the electrolyte, the reaction between the cathode and the binder, and the reaction between the anode and the electrolyte, electrolyte decomposition reaction. However, due to the difference in the battery material system, the order of the latter two reactions will be different. The reaction between the cathode and the binder is neglected due to the weak reaction. In this work, a heating rod is used to detonate battery A to achieve thermal runaway propagation, so the radiant heat of the heating rod is ignored for batteries B, C, and D. Given the above, two primary heating mechanisms are considered in the heat transfer model, that is, internal heating caused by exothermic reactions and external heating caused by the thermal runaway propagation. The internal heating caused by exothermic reactions is shown in Eq. 2:24,25
| \begin{equation} Q_{\text{chem}} = Q_{\text{sei}} + Q_{\text{ne}} + Q_{\text{pe}} + Q_{\text{e}} \end{equation} | (2) |
| \begin{equation} Q_{\text{sei}} = H_{\text{sei}} \cdot W_{\text{c}} \cdot A_{\text{sei}} \cdot \exp \left(-\frac{E_{\text{a,sei}}}{RT}\right) \cdot \text{c}_{\text{sei}}^{m_{\text{sei}}} \end{equation} | (3) |
| \begin{equation} Q_{\text{ne}} = H_{\text{ne}} \cdot W_{\text{c}} \cdot A_{\text{ne}} \cdot \exp \left(-\frac{t_{\text{sei}}}{t_{\text{sei,ref}}} \right) \cdot c_{\text{ne}}^{m_{\text{ne}}} \cdot \exp \left(-\frac{E_{\text{a,ne}}}{RT}\right) \end{equation} | (4) |
| \begin{equation} Q_{\text{pe}} = H_{\text{pe}} \cdot W_{\text{p}} \cdot A_{\text{pe}} \cdot \alpha^{m_{\text{pe1}}} \cdot (1 - \alpha)^{m_{\text{pe2}}} \cdot \exp \left(-\frac{E_{\text{a,pe}}}{RT}\right) \end{equation} | (5) |
| \begin{equation} Q_{\text{e}} = H_{\text{e}} \cdot W_{\text{e}} \cdot A_{\text{e}} \cdot c_{\text{e}}^{m_{\text{e}}} \cdot \exp \left(-\frac{E_{\text{a,e}}}{RT}\right) \end{equation} | (6) |
| \begin{equation} Q_{\text{r}} = \sigma T_{\text{f}}^{4} [1 - \exp (-kL_{\text{m}})] \end{equation} | (7) |
| \begin{equation} Q_{\text{cond, battery}} = \lambda A \left( \frac{T_{1} - T_{2}}{w} \right) \end{equation} | (8) |
| \begin{equation} Q_{\text{cond, air}} = \cfrac{2\pi LK_{\text{air}} (T_{1} - T_{2})}{\cosh^{-1} \biggl(\cfrac{4w^{2} - D_{1}^{2} - D_{2}^{2}}{2D_{1}D_{2}} \biggr)} \end{equation} | (9) |
| \begin{equation} Q_{\text{rad}} = \cfrac{\sigma (T_{1}^{4} - T_{2}^{4})}{\cfrac{1 - \varepsilon_{1}}{A_{1}\varepsilon_{1}} + \cfrac{1}{A_{1}F_{12}} + \cfrac{1 - \varepsilon_{2}}{A_{2}{\varepsilon_{2}}}} \end{equation} | (10) |
| \begin{equation} F_{12} = \frac{1}{2\pi}\left[\pi + \sqrt{\left(1 + \frac{w}{r} \right)^{2} {}- {4}} - \left(1 + \frac{w}{r} - 2\cos^{-1}\left(\frac{2r}{r + w} \right)\right)\right] \end{equation} | (11) |
| \begin{equation} Q_{\text{ejecia}} = C_{\text{p}}m_{\text{ejecta}}(T_{\text{ejecta}} - T_{\text{air}}) \end{equation} | (12) |
| \begin{align} \rho C_{\text{p}} \frac{\partial T}{\partial t} &= \vec{\nabla}(\lambda \cdot \vec{\nabla} T) + Q_{\text{chem}} - Q_{\text{cond, battery}} - Q_{\text{cond, air}} \notag\\ &\quad- Q_{\text{rad}} - Q_{\text{ejecta}} \end{align} | (13) |
| \begin{align} \rho C_{\text{p}} \frac{\partial T}{\partial t} &= \vec{\nabla} (\lambda \cdot \vec{\nabla} T) + Q_{\text{chem}} + Q_{\text{r}} - Q_{\text{cond, battery}} \notag\\ &\quad- Q_{\text{cond, air}} - Q_{\text{rad}} - Q_{\text{ejecta}} \end{align} | (14) |
When the thermal runaway has happened in battery A, the heat generated by the explosion is directly transferred to the adjacent battery, causing the heating up. The heat will continue to transfer from high to low through thermodynamic movement. In addition, there is also a thermal effect on surrounding batteries caused by the flame which is generated by the thermal runaway of lithium-ion batteries with 100 % SOC. Eventually, the heat accumulates to a threshold, thermal runaway spreads within the battery pack, and batteries B, C, and D explode one by one.
3.2 Analysis of temperature change and the time of thermal runawayFigure 3 shows the temperature variation of lithium-ion batteries during the whole experiment. The curves can be divided into three parts: (I) the steady increase stage, (II) the sudden change stage, and (III) cooling down the stage. During the stage of steady increase, the temperature of the battery increases mainly caused by a heating rod and only a little heat from the self-reactions in the battery. While in the sudden change stage, the heat of self-reaction of materials in the battery increases greatly and massive uncontrolled exothermic reactions are triggered and lead to the release of a large number of gases, sparks, and heat. Consequently, the temperature is elevated to 500 °C in a lithium-ion battery with 30 % SOC and 750 °C in a lithium-ion battery with 100 % SOC in a few seconds.
Temperature variation curves for lithium-ion batteries with aerogel felt of different thicknesses: the red curve represents the battery with 30 % SOC, and the black curve represents the battery with 100 % SOC. Battery A to battery D is distinguished from different shapes. Figure 3a is the temperature variation curve of lithium battery under the barrier thickness of 1 mm. Figure 3b is the temperature variation curve of lithium battery under the barrier thickness of 3 mm. Figure 3c is the temperature variation curve of lithium battery under the barrier thickness of 6 mm.
From the curve shown in Fig. 3, the curve of a lithium-ion battery with 100 % SOC is higher than that of a lithium-ion battery with 30 % SOC. This is mainly ascribed to the SOC, the larger the SOC, the higher the maximum temperature. And the heating time of battery A becomes longer when the thickness of aerogel felt is increased from 1 mm to 6 mm, the reason for this phenomenon is that the heat propagation to the second battery for thermal runaway becomes slower with the increase of the thickness. In addition, the difference in the maximum temperature is small with the increase in thickness. On the one hand, this is an attribute of the heat emission being reduced by an increase in thickness. On the other hand, there is little effect on the internal materials of lithium-ion batteries by the propagation of thermal runaway. Thereby, comparing the difference in the temperature of the three temperature curves, it is not difficult to see that there is a small difference in the maximum temperature of the battery pack, but the thermal runaway time of the battery pack is delayed with the increase of the thickness.
The temperature of the battery is collected by a paperless recorder every 1 second. And the temperature change rate curve is obtained by taking the derivative of the temperature variation curve, the formula is shown in (15).
| \begin{equation} \frac{\text{d}T}{\text{d}t} = \frac{T_{i+1} - T_{i}}{t_{i + 1} - t_{i}},\quad (i = 1,2,3 \ldots n) \end{equation} | (15) |
Figure 4 depicts the temperature change rate curve of lithium-ion batteries under different thicknesses of aerogel felt. There are four peaks in the curve, corresponding to the thermal runaway behavior of four lithium-ion batteries respectively. The average temperature change rate of the battery decreases from 118.5 °C s−1, 96.9 °C s−1, 101.2 °C s−1 to 39.4 °C s−1, 26.4 °C s−1, 37.7 °C s−1 at the thickness of 1 mm, 3 mm and 6 mm, with the decrease of SOC of lithium-ion battery from 100 % to 30 %. And the triggering time of thermal runaway is delayed. This is ascribed to the heat generated by reaction with electrolyte due to different lithium carbon embedded in the cathode of lithium-ion batteries with 30 % SOC and 100 % SOC, resulting in different response times of thermal runaway of the battery. Meanwhile, there is a negative value of temperature change rate in the process of thermal runaway, indicating that the temperature decreases instantaneously in a short time. And the main reason for this phenomenon is the release of gas and the ejection of some active substances during the opening of the safety valve in the thermal runaway, resulting in a short-term decrease in the battery temperature. Specifically, under the 30 % SOC of battery, battery C jumped to the maximum heating rate of 46.5 °C s−1 at 649 s in the thickness of 1 mm barrier. In the thickness of 3 mm barrier, battery B jumped to the maximum heating rate of 28.1 °C s−1 at 949 s. In the thickness of 6 mm barrier, battery B jumped to the maximum heating rate of 46 °C s−1 at 1415 s. Under the 100 % SOC of the battery, battery C jumped to the maximum heating rate of 173.5 °C s−1 at 502 s in the thickness of 1 mm barrier. In the thickness of 3 mm barrier, battery B jumped to a maximum temperature rise rate of 130.9 °C s−1 at 574 s. In the thickness of 6 mm barrier, battery C jumped to a maximum temperature rise rate of 109.3 °C s−1 at 1047 s. It can be seen that the location of the maximum heating rate is random, which indicates that the influence of aerogel felt thickness on the maximum heating rate is small.
The temperature change rate of lithium-ion batteries with aerogel felt of thickness at 1 mm, 3 mm, and 6 mm: the figures on the left are the temperature change rate curve of lithium-ion batteries with 30 % SOC, and the figures on the right are the temperature change rate curve of lithium-ion batteries with 100 % SOC. Battery A to battery D are distinguished from the different color curves. Figure 4a is the temperature change rate curve of lithium battery under the barrier thickness of 1 mm. Figure 4b is the temperature change rate curve of lithium battery under the barrier thickness of 3 mm. Figure 4c is the temperature change rate curve of lithium battery under the barrier thickness of 6 mm.
In addition, as an important parameter to evaluate the barrier effect of the thickness of aerogel felt on thermal runaway propagation of lithium-ion battery, thermal runaway triggering time and propagation time were recorded and analyzed in detail. They are defined by Eqs. 16 and 17.30 And the calculated results are shown in Figs. 5 and 6 respectively.
| \begin{align} t_{\text{onset},i} = \min \left\{ k:\left[ \frac{T_{i}(k + 1) - T_{i}(k)}{t_{i}(k + 1) - t_{i} (k)} > 30\,{{}^{\circ}\text{C}}\,\text{s$^{-1}$}\right] \right\},&\notag\\ (i = 1,2,3,4)& \end{align} | (16) |
| \begin{equation} t_{i,i + 1} = t_{\text{onset},i + 1} - t_{\text{onset},i}, \quad (i = 1,2,3,4) \end{equation} | (17) |
Mean triggering time of lithium-ion battery with the thermal runaway in thickness at 1 mm, 3 mm and 6 mm: the figure is the error bar of lithium-ion battery triggering time of thermal runaway. The black curve represents the lithium-ion battery with 30 % SOC, the red curve represents the lithium-ion battery with 100 % SOC. Battery A to battery D is distinguished from different shapes.
The thermal runaway propagation time for lithium-ion batteries with aerogel felt of different thicknesses at 1 mm, 3 mm and 6 mm: the figure is the error bar of the thermal runaway propagation time of the lithium-ion battery. Where i denotes the order of the battery, 1, 2, 3 and 4 represent batteries A, B, C and D respectively. ti,i+1 represents the thermal runaway propagation time between the batteries. The black curve represents the lithium-ion battery with 30 % SOC, the red curve represents the lithium-ion battery with 100 % SOC. Battery A to battery D is distinguished from different shapes. Figure 6a represents the thermal runaway from battery A to battery B and C, and Fig. 6b represents the thermal runaway from battery B and C to battery D.
Figure 5 shows the mean triggering time of lithium-ion batteries with thermal runaway. It is evident that the tonset,i in lithium-ion batteries delays as the increase of the thickness of the aerogel felt. Specifically, for the farthest battery D with 30 % SOC, the tonset,D are 929 s, 1216 s, and 1942 s for the thickness of 1 mm, 3 mm, and 6 mm barrier, respectively, while the tonset,D for battery D with 100 % SOC are 546 s, 709 s, and 1210 s, respectively. This shows that the explosion intensity of the battery with different charges leads to different thermal runaway propagation times, but both tonset,i of two kinds of charge are delayed with the increase of thickness. It can be seen from Fig. 2 that there is the same distance between battery A and battery B, and C, which makes the tonset,i of battery B and battery C close. In addition, only the tonset,i of battery A is almost in the same position in Fig. 5. It may account for the direct contact with the heating rod.
Figure 6 shows the mean thermal runaway propagation time for lithium-ion batteries during thermal runaway. The ti,i+1 from battery A to battery B, C, are shown in Part (a), and the ti,i+1 from battery B, C to battery D are demonstrated in part (b). It is worth noting that the ti,i+1 rises with the increase of the thickness of aerogel felt, which corresponds to the variation of the tonset,i. This demonstrates that the ti,i+1 and tonset,i are effectively delayed with the increase of the barrier thickness. In Fig. 6, the ti,i+1 of part (a) is longer than that of part (b). Specifically, when the battery temperature is close to thermal runaway, the heating rate of the lithium-ion battery increases in a short time, so that the battery loses less energy and more energy obtained by the next battery. And thus thermal runaway time required for a thermal runaway is shorter.
3.3 Analysis of gas concentration and toxicityUnder the thickness of 1 mm, 3 mm and 6 mm barrier, the O2 consumption and COx production for lithium-ion batteries with 30 % SOC and 100 % SOC are shown in Figs. 7 and 8, respectively.
O2 consumption for lithium-ion batteries of different charges at the thickness of 1 mm, 3 mm and 6 mm: the black curve represents the lithium-ion battery with 30 % SOC, and the red curve represents the lithium-ion battery with 100 % SOC. And the thickness of aerogel felts is distinguished by different shapes. The time difference in the figure represents the length of stage II of stabilization.
CO and CO2 concentration for lithium-ion batteries of different charges at the thickness of 1 mm, 3 mm and 6 mm: the left axis represents the CO and CO2 concentration of a lithium-ion battery with 100 % SOC, and the right axis represents the CO and CO2 concentration of a lithium-ion battery with 30 % SOC. The CO and CO2 concentration is distinguished by a different color. Figure 8a is the CO and CO2 concentration curve of lithium battery under the barrier thickness of 1 mm. Figure 8b is the CO and CO2 concentration curve of lithium battery under the barrier thickness of 3 mm. Figure 8c is the CO and CO2 concentration curve of lithium battery under the barrier thickness of 6 mm.
As seen from the figures, all the curves of O2 consumption decrease step by step, CO and CO2 production curves increase gradually. There are 3 main stages of the reaction process. Stage I, lithium-ion battery is smoldered, no O2 is consumed, and CO2 generation is small. Stage II, a rupture of the battery’s safety valve causes O2 to react with the electrolyte to form the CO2. The decomposition reaction of the sub-stable layer of the SEI membrane produces CO2. And the combustible gas generated by the decomposition of the internal material reacts with air to produce CO2. Then the CO2 concentration rises to a small peak and so does the O2 consumption. And the increase in the battery’s temperature leads to the release of O2 from the pyrolysis of anode and cathode materials, and a small increase in O2 concentration, while the weakening of the flame leads to a small decrease in CO2 concentration. In stage III, under the dual action of external radiation heat and the battery’s own thermal feedback, the pyrolysis reaction of the battery’s internal materials accelerates and generates more CO2. And the accumulation of heat leads to the thermal runaway of batteries. Violent combustion of batteries causes the O2 concentration to drop sharply to the minimum and the CO2 and CO concentrations to rise sharply to the maximum. Then followed by a weakening flame, a slow rise in O2 and a gradual decrease in CO2 and CO concentrations.
In Fig. 7, the increase in the thickness of aerogel felt in the battery pack resulted in the prolonged O2 consumption time of stage II. Specifically, under the thickness of 6 mm barrier, the O2 consumption times are up to 805 s in the battery with 30 % SOC and 993 s in the battery with 100 % SOC. Under the 30 % SOC of battery, O2 consumption in the thickness of 1 mm, 3 mm and 6 mm barrier is 3.1 %, 1.82 % and 4.76 % respectively. Under the 100 % SOC of battery, O2 consumption in the thickness of 1 mm, 3 mm and 6 mm barrier is 10.1 %, 8.1 % and 7 % respectively. And O2 consumption in the battery with 100 % SOC is higher than that in the battery with 30 % SOC. Overall, it seems that the increase of the thickness can effectively delay the O2 consumption time and buy some time for fire rescue. And the O2 consumption reaches the maximum at 6 mm and 1 mm in different SOC, respectively, which shows that the increase in thickness and the change of SOC are created a little effect on the consumption of O2.
In Fig. 8, the trend of CO and CO2 concentration variation curves is the same. There are two obvious peaks in the curve of the battery with 100 % SOC. While the curve of the battery with 30 % SOC shows piecewise change, and a few curves show ladder-like continuous change. Furthermore, it is obvious that the generation time of CO2 and CO in stage II becomes longer with the increase in thickness. Specifically, under the thickness of 6 mm barrier, CO2 and CO generation times are 761 s and 831 s in the battery with 30 % SOC, 1085 s and 1128 s in the battery with 100 % SOC. And the CO2 and CO concentrations of battery with 30 % SOC reach to a maximum of 4.1 % and 1.2 % at 6 mm and 3 mm barrier, respectively, and the CO2 and CO concentrations of battery with 100 % SOC reach a maximum of 8.8 % and 3 % at 3 mm barrier, respectively. These data showed that due to more obvious fire released by the lithium-ion battery with 100 % SOC, the toxic and flammable gases are relatively small, and the generation time of CO and CO2 can be effectively delayed with the increase of barrier thickness. In addition, under the same SOC, the difference between CO2 and CO production is small, but the concentrations of CO2 and CO are related to the charge of the lithium-ion battery itself.
In order to analyze the toxicity of gases released from lithium-ion batteries, a simplified model for quantitative assessment of N-GAS toxicity is established,29 and it is shown in Eq. 18.
| \begin{equation} \mathrm{TOX}_{\text{N-GAS}} = \frac{m\varphi (\text{CO})}{\varphi (\text{CO$_{2}$}) - b} + \frac{21 - \varphi (\text{O$_{2}$})}{21 - LC_{50}(\text{O$_{2}$})} \end{equation} | (18) |
| SOC | Parameter | O2/% | CO2/% | CO/% | N-GAS |
|---|---|---|---|---|---|
| 30 %SOC | 1 mm | 3.1 | 3 | 0.7 | 1.0011 |
| 3 mm | 1.82 | 3.1 | 1.2 | 1.0017 | |
| 6 mm | 4.76 | 4.1 | 0.7 | 1.0003 | |
| 100 %SOC | 1 mm | 10.1 | 8.7 | 2.8 | 0.9978 |
| 3 mm | 8.1 | 8.8 | 3 | 0.9987 | |
| 6 mm | 7 | 8.1 | 2.78 | 0.9992 |
It can be seen that gas released in the lithium-ion batteries with 30 % SOC is more than that in the battery with 100 % SOC, and the toxicity is all higher than 1. This is attributed to the long heating process under low SOC, resulting in a relatively long thermal runaway process. Under the battery with 30 % SOC, 1.0003 at the thickness of 6 mm barrier is the lowest, indicating that the released gas concentration can kill some animals. The N-GAS in the battery of 100 % SOC is all between 0.8 and 1, with the lowest being 0.9987 at the thickness of 1 mm barrier. Overall, it shows that there is little connection between the gas toxicity and the increase of thickness of aerogel felt, but it’s related to its SOC.
3.4 Analysis of mass lossThe mass loss mainly includes the loss of cartons, the solid particles, liquid and gas ejected from the battery after heating. The equation of dimensionless mass loss is established to reveal the mass loss law of lithium-ion batteries and is shown in (19).31
| \begin{equation} W^{*} = \frac{W_{t} - W}{W_{t} - W_{a}} \end{equation} | (19) |
Dimensionless mass loss for lithium-ion batteries of different charges at the thickness of 1 mm, 3 mm and 6 mm 1 represents that there is only one battery with thermal runaway caused mass loss, 2 represents that there are two batteries with thermal runaway caused mass loss, 3 represents that there are three batteries with thermal runaway caused the mass loss, 4 represents that there are four batteries with thermal runaway caused the mass loss.(13,)
As seen in Fig. 9, the lithium-ion battery with 30 % SOC only released gas during the whole process, making the curve gentle. While the reaction of thermal runaway in a lithium-ion battery with 100 % SOC is violent, which will exert a reaction force on the electronic scale, resulting in more obvious fluctuations in the curve. The W$^{*}$ decreased in the same period with the increase of the thickness, and at the end time of W$^{*}$ of the battery with 100 % SOC was faster than that of 30 % SOC. This indicates that due to the higher charge, the heating rod causes more material to eject and ignite faster. And under the thickness of 6 mm barrier, the mass loss time of thermal runaway of the second lithium-ion battery was 1042 s in the battery with 30 % SOC and 1255 s in the battery with 100 % SOC. It can be seen from the data that the mass loss time is delayed and the mass loss rate is reduced with the increase of aerogel felt thickness.
This work was partly supported by the key project of National Natural Science Foundation of China (U2033206) and the R&D program of Sichuan Provincial Science and Technology Department (2018GZYZF0069).
Quanyi Liu: Project administration (Lead)
Qian Zhu: Investigation (Equal), Writing – original draft (Lead), Writing – review & editing (Equal)
Wentian Zhu: Investigation (Equal), Writing – review & editing (Equal)
Xiaoying Yi: Investigation (Equal), Writing – review & editing (Equal)
The authors declare no conflict of interest in the manuscript.
National Natural Science Foundation of China Key Project: U2033206
Sichuan Provincial Science and Technology Department R&D Program Project [2018GZYZF0069].
The content of this paper has been published by Xiaoying Yi as a Ph.D. thesis at the Civil Aviation Flight University of China in 2022. URL: https://doi.org/10.2722/d.cnki.gzgmh.2021.000088.
