Advances in the Battery Thermal Management Systems of Electric Vehicles for Thermal Runaway Prevention and Suppression

Abstract

In response to the global imperative to reduce greenhouse gas emissions and fossil fuel dependency, electric vehicles (EVs) have emerged as a sustainable transportation alternative, primarily utilizing lithium-ion batteries (LIBs) due to their high energy density and efficiency. However, LIBs are highly sensitive to temperature fluctuations, significantly affecting their performance, lifespan, and safety. One of the most critical threats to the safe operation of LIBs is thermal runaway (TR), an uncontrollable exothermic process that can lead to catastrophic failure under abusive conditions. Moreover, thermal runaway propagation (TRP) can rapidly spread failures across battery cells, intensifying safety threats. To address these challenges, developing advanced battery thermal management systems (BTMS) is essential to ensure optimal temperature control and suppress TR and TRP within LIB modules. This review systematically evaluates advanced cooling strategies, including indirect liquid cooling, water mist cooling, immersion cooling, phase change material (PCM) cooling, and hybrid cooling based on the latest studies published between 2020 and 2025. The review highlights their mechanisms, effectiveness, and practical considerations for preventing TR initiation and suppressing TRP in battery modules. Finally, key findings and future directions for designing next-generation BTMS are proposed, contributing valuable insights for enhancing the safety and reliability of LIB applications.

1. Introduction

In recent decades, environmental pollution and global warming have become important challenges for societies around the world. Many human activities that have greenhouse gas emissions have contributed significantly to climate change [1]. One of the largest sources of these emissions is the transportation sector, where fossil fuel-powered internal combustion engine (ICE) vehicles are a major contributor to air pollution and carbon dioxide emissions [2,3]. In response to growing environmental concerns, electric vehicles (EVs) have emerged as a promising alternative to traditional ICE vehicles, offering a cleaner and more sustainable transportation option [4,5].

Lithium-ion batteries (LIBs) are the core components of EVs, serving as their primary energy storage systems. LIBs are favored for their high energy density, long lifecycle, low self-discharge rate, and high conversion rate, making them ideal for use in EVs [6,7,8]. However, the performance, lifetime, and safety of LIBs are significantly dependent on their operating temperature range. Studies have shown that LIBs perform best within a temperature range of 25–40 °C [9,10,11]. In addition, the temperature difference between cells in the battery module should be kept below 5 °C to ensure uniform performance and prevent rapid aging [12,13,14]. Operation at temperatures outside this range can have adverse effects. Specifically, LIBs operating at low temperatures reduce battery capacity and accelerate aging, while high temperatures can lead to loss of performance and loss of thermal control, posing significant safety risks in extreme cases [15,16,17].

Additionally, thermal runaway (TR) in LIBs is a significant safety concern, primarily triggered by abusive conditions such as thermal, electrical, and mechanical stress [18,19,20]. As the temperature of a battery cell increases, exothermic reactions within the cell become more intense, leading to an uncontrollable rise in temperature. Once a critical temperature threshold is surpassed, typically between 130 and 200 °C, the battery can enter TR, where the temperature can escalate to 600–1000 °C, often accompanied by the emission of flammable gases, smoldering, and combustion [21,22]. The heat generated from a single cell undergoing TR can appear limited. However, in densely packed battery systems, the proximity of cells can result in thermal runaway propagation (TRP). This chain reaction occurs when the heat from a failing cell triggers TR in adjacent cells, causing a rapid spread of TR throughout the battery pack. The intense heat generated by TRP can lead to severe consequences, including widespread fires, explosions, and damage to surrounding systems, posing a significant threat to public safety [23]. Notable fire incidents, such as those involving EVs, such as the BMW i3, Tesla Model X, and Porsche Panamera, highlight the danger of TRP, where a single cell failure triggered catastrophic fires [22]. The risk of TR and TRP is exacerbated by common causes such as overheating, internal short circuits (ISC), overcharging, and manufacturing defects, which initiate the internal exothermic reactions within the cell. Furthermore, the spread of TR can be accelerated by the design of the battery, with heat transferring rapidly through the battery shell to neighboring cells. Consequently, effective battery thermal management systems (BTMS) are critical in preventing the onset of TR and suppressing its propagation, ensuring the safety and reliability of the LIB application systems [24,25].

Liquid cooling and phase change material (PCM) cooling are two advanced battery thermal management techniques that play a crucial role in preventing and suppressing TR in LIBs. Liquid cooling systems are widely recognized for their high heat transfer efficiency as they utilize a circulating coolant to absorb and dissipate heat from the battery cells. This method allows for precise temperature control and uniform cooling, which is vital for maintaining the battery within the optimal operating temperature range. Liquid cooling also offers the flexibility to scale for various battery sizes and configurations, making it suitable for applications ranging from EVs to large-scale energy storage systems [26,27]. On the other hand, PCM cooling systems provide the solution to manage battery temperature effectively by utilizing materials that absorb and store excess heat during phase transitions from solid to liquid. PCM cooling can effectively reduce the risk of TR by preventing overheating and maintaining the battery’s temperature within safe limits. Additionally, the large latent heat of the PCM can significantly delay the TRP time from the TR-triggered cell to other cells in the battery pack [28,29,30,31]. When integrated into battery packs, both liquid cooling and PCM cooling systems offer outstanding benefits, ensuring thermal stability, enhancing battery lifespan, and improving overall safety by reducing the likelihood of TR incidents.

Recently, several review articles have been published addressing thermal management strategies to prevent TR in LIBs, as summarized in

Table 1. Recent reviews on battery thermal management strategies to prevent TR in LIBs.

Review	Focus of Review	Highlights	Missing Topics/Future Perspectives
Jiang et al. (2022) [32]	Comprehensive review of BTMS for LIBs across normal, sub-zero, and abusive conditions. Discusses battery cooling (air cooling, liquid cooling, PCM cooling, and hybrid cooling) and battery heating methods. Covers TR origins and prevention strategies, but is not exclusively focused on TR prevention, and addresses overall thermal management, including general operation, low-temperature performance, and abusive conditions.	Reviews heat generation mechanisms during both normal use and abusive events. Summarizes BTMS methods, including air cooling, liquid cooling, PCM cooling, and hybrid cooling. Discusses battery heating methods such as alternating current (AC) heating, self-heating, and external heating under sub-zero temperatures. Addresses TR prevention via both battery material modification and thermal management design. Highlights research challenges, especially in developing BTMS capable of handling wide temperature ranges and multiple abusive scenarios.	Does not specifically address thermal runaway propagation (TRP) suppression. Focuses on both traditional cooling, such as air cooling. Lacking in-depth analysis and evaluation of advanced cooling strategies, namely liquid cooling and PCM cooling. Omits recent developments in advanced BTMS for TR prevention and TRP suppression.
Shahid et al. (2022) [19]	Comprehensive review of the entire TR process in LIBs, including initiation, propagation, gas venting, prediction, prevention, and mitigation. Covers both BTMS for TR prevention and mitigation strategies (insulation, fire barriers, and fire suppression) after TR occurs. BTMS is an important part, but not the sole focus.	Summarizes thermal abuse, mechanical abuse, and electrical abuse mechanisms that cause TR. Details experimental and numerical studies on TRP and gas emission during TR events. Reviews BTMS technologies (air cooling, liquid cooling, PCM cooling, and hybrid cooling systems) as key preventive measures. Critically discusses mitigation strategies post-TR, including fire suppression, explosion barriers, and thermal insulation methods. Provides future research directions, including the development of better TR prediction models, lightweight BTMS, and integrated safety designs for EV batteries.	Provides a limited performance evaluation of advanced cooling methods specifically for EV battery modules. Insufficient coverage of recent advancements in BTMS for TR prevention and TRP suppression. Lacks clear identification of key future research directions for the development of advanced BTMS.
Chavan et al. (2023) [20]	Reviews thermal management systems for LIBs with an emphasis on TR mitigation using battery cooling methods. Focused primarily on cooling technologies (air cooling, liquid cooling, PCM cooling, heat pipes, and hybrid cooling) as a strategy to prevent and suppress TR in EV batteries. Discusses TR mechanisms, TR modeling, and TRP prevention techniques.	Provides a detailed classification and evaluation of active and passive battery cooling techniques specifically for TR prevention. Highlights the importance of hybrid cooling (PCM and liquid cooling) to balance normal operation efficiency and TR suppression capability. Reviews the necessity of thermal insulation layers between cells to enhance TR propagation resistance. Stresses that cooling alone is insufficient for complete TR suppression, recommending combined strategies (cooling and insulation). Suggests future research directions in improving thermal conductivity, enhancing system compactness, and integration with battery management systems (BMS).	Limited in-depth analysis of advanced battery thermal management strategies specifically aimed at preventing TRP. Does not incorporate the most recent advancements in BTMS published in the past few years.
Zhi et al. (2024) [33]	Comprehensive review of technologies for the prevention, early detection, and mitigation of TR in LIBs. Covers intrinsic safety technologies (material modifications), thermal management systems (air, liquid, PCM, and hybrid cooling), early warning and detection methods, and fire prevention/extinguishing techniques. Focuses on TR broadly, not specifically on BTMS alone.	Summarizes thermal management technologies (air, liquid, PCM, and hybrid cooling) as part of TR prevention strategies, but thermal management is not the sole focus. Includes extensive discussion on material innovation (flame-retardant electrolytes, electrodes, separators) and early warning technologies (sensor systems). Reviews fire prevention and fire suppression technologies of TR. Proposes challenges and future directions for comprehensive TR risk mitigation rather than focusing exclusively on advanced BTMS designs.	While the review includes an evaluation of conventional cooling methods, it lacks a focused and in-depth analysis of advanced BTMS tailored for LIB applications. Insufficient research has addressed the effectiveness of advanced cooling technologies in suppressing TRP, limiting the practical insights for improving battery safety. The review does not provide specific or detailed recommendations for the future development of advanced BTMS aimed at preventing TR and suppressing TRP, which are critical for high-energy battery applications.
Yang et al. (2024) [34]	Reviews TR mechanisms and thermal management technologies specifically for LIBs in electric aircraft. Focuses on TR prevention and mitigation strategies tailored for aviation applications. Covers both enhanced heat removal techniques (air cooling, liquid cooling, and PCM cooling) and heat/fire propagation prevention methods (thermal insulation, fire-retardant technologies).	Evaluates various thermal management strategies (air cooling, liquid cooling, PCM cooling, and thermochemical cooling) for their suitability in electric aircraft, based on weight, power, and effectiveness. Highlights the potential of PCM combined with thermochemical materials and bio-inspired insulation materials as promising lightweight solutions. Recommends passive fire-retardant methods combined with PCM.	The review primarily focuses on advanced BTMS for TR prevention in aviation applications, limiting its relevance to EV contexts. It lacks a detailed and systematic evaluation of advanced BTMS strategies specifically designed to prevent TRP in LIB systems. There is an absence of clearly defined recommendations for the future development and optimization of advanced BTMS aimed at enhancing both TR prevention and TRP suppression under realistic operating conditions.

. Unlike previous reviews, this study provides an in-depth and systematic analysis on thermal runaway prevention/suppression using advanced cooling strategies, including indirect liquid cooling, water mist cooling, immersion cooling, PCM cooling, and hybrid cooling systems. The novelty of this review distinctly lies in its comprehensive and specific focus on evaluating advanced BTMS tailored explicitly for both the TR prevention and TRP suppression in LIB modules used in EVs, while most previous reviews have been limited to a TR prevention perspective. Additionally, this review uniquely incorporates and synthesizes the most recent advancements published between 2020 and 2025, addressing the significant gap identified in the prior literature that frequently overlooked the latest developments or offered limited insights into advanced BTMS for TR prevention and TRP suppression. Furthermore, this review explicitly discusses the effectiveness and limitations of each advanced cooling approach, offering clear evaluations and critical comparisons that were insufficiently addressed in earlier studies. Thus, the present review distinctly advances the current state-of-the-art by providing practical insights and specific directions to guide ongoing and future research in battery safety and thermal management. The structure of this review is as follows: Section 2 outlines the mechanisms leading to TR and the severe impacts of the heat generated during TR events; Section 3 discusses advanced cooling strategies for TR prevention and TRP suppression in LIBs; and Section 4 summarizes the key findings and presents recommendations for future research directions.

2. TR Characterization

2.1. TR Mechanism

The basic mechanisms that lead to TR in LIBs include mechanical abuse, electrical abuse, and thermal abuse [18]. These mechanisms can initiate a series of chain reactions that escalate the temperature within the battery and propagate failures, potentially causing fires or explosions. The abusive conditions associated with the TR mechanism of LIBs are shown in Figure 1.

Mechanical abuse typically arises from impacts such as vehicle crashes or external forces that deform the battery pack. Such deformation can lead to the rupture of the separator between the anode and cathode, which, in turn, causes an ISC [19]. The ISC generates significant heat and leads to the initiation of TR. Additionally, penetration by an external object (e.g., a nail) can trigger a short circuit and lead to rapid heat generation as the battery discharges. The extent of damage depends on the location of the penetration. In particular, low heat dissipation areas are more susceptible to TR [20]. As the mechanical forces increase, the battery pack’s deformation can become non-uniform, concentrating stress and elevating the risk of failure.

Electrical abuse in LIBs primarily involves external short circuits (ESC) and overcharging or overdischarging. ESC occurs when electrodes with different voltages come into contact, often due to external forces such as water immersion, contamination, deformation during mechanical abuse, or electric shock [18]. When an ESC occurs, the current increases rapidly to its peak value and then decreases to a plateau before dissipating as the battery discharges. This process generates heat within the circuit. However, in typical ESC conditions, this heat is often insufficient to trigger TR unless certain conditions are met, such as an increase in the mass transfer coefficient for lithium ions at the anode or a larger surface area of the anode, which leads to greater current flow and faster heat generation [35]. Overcharging is another form of electrical abuse and a significant contributor to TR. This condition arises when the battery is charged beyond its recommended voltage limit, typically due to a failure in the battery management system (BMS) that prevents the battery from stopping the charging process [18]. Overcharging results in excessive energy being introduced into the battery, leading to increased heat generation from both ohmic heating and side reactions at the electrodes [35,36]. These reactions are particularly severe at the anode, where lithium plating can occur due to excessive lithium intercalation, causing an increase in temperature. In addition, overcharging can lead to the collapse of the cathode material structure and the release of oxygen, which further accelerates electrolyte decomposition and produces additional gases [37,38]. Similarly, overdischarging occurs when the battery is drained beyond its capacity, causing a negative voltage in certain cells, leading to abnormal heat generation and potential capacity degradation [18]. Overdischarging can also cause the dissolution of copper collectors in the battery, resulting in ISC and TR [39,40,41].

Thermal abuse occurs when the battery experiences overheating, often due to mechanical or electrical failures, poor battery connections, and external fires. Thermal abuse can escalate quickly, as the rising internal temperature accelerates the breakdown of the separator and the electrolyte, leading to ISC and further increasing heat generation. This can create a series of chain reactions where the temperature rise causes additional breakdowns, culminating in full-scale TR [19,20,42].

2.2. TRP

TRP refers to the spread of the TR phenomenon from the triggered TR cell to neighboring cells within a battery pack. Once TR is initiated in a single cell, it generates significant heat that can trigger the same event in adjacent cells and potentially cause a TRP throughout the entire pack [19]. Feng et al. conducted an experimental investigation on penetration-induced TRP within a large-format LIB pack. In the six-battery module test, the study indicated that a maximum temperature up to 791.8 °C occurred within a single battery during TRP. The analysis showed that approximately 12% of the total heat released by the initiating battery is sufficient to trigger TR in adjacent cells, with the majority of heat transferring through the battery shell rather than the pole connector [43]. These temperatures and heat transfer characteristics were recorded without active cooling and represent the severity of uncontrolled TRP in densely packed battery modules. Chen et al. explored TRP in cylindrical battery packs and found that larger battery packs exhibited more serious TRP. Specifically, the study showed that 32 battery cells simultaneously ignited during the TR event in a battery pack consisting of 99 cells [44]. The intensity and speed of TRP are also influenced by the state of charge (SOC) of the batteries with higher SOCs, leading to faster and more intense TRP [45].

2.3. Heat Generation During TR

During TR in LIBs, a series of exothermic reactions generate significant amounts of heat, rapidly escalating the internal temperature of the battery. The heat generation occurs in multiple stages, each corresponding to specific reactions triggered by elevated temperatures. These reactions include the decomposition of the solid electrolyte interphase (SEI), reactions at the negative and positive electrode–electrolyte interfaces, and the decomposition of the electrolyte. Additionally, heat can also be generated due to an ISC, which occurs when the separator partially melts as the temperature rises, allowing for the anode and cathode materials to come into direct contact. This ISC further contributes to heat release. The total heat generation is calculated by summing the heat produced by each reaction, as defined by the SEI decomposition reaction ($Q_{sei}$), negative-electrolyte reaction ($Q_{ne}$), positive-electrolyte reaction ($Q_{pe}$), electrolyte decomposition ($Q_e$), and the heat generated by ISC ($Q_{ISC}$). The total heat generation during TR is calculated using Equation (1) [46],

$$Q_{TR}=Q_{sei}+Q_{ne}+Q_{pe}+Q_e+Q_{ISC}$$

(1)

The ranges of the starting reaction temperatures and the formulas for calculating the heat released for each reaction process are presented in

Table 2. Heat generation equations of TR reactions [22,46,47].

TR Reactions	Temperature Range for the Initiation of Reaction	Reaction Rate Equation	Heat Generation Equation
SEI decomposition reaction	60–130 °C	$R_{sei}={\displaystyle \frac{d{c}_{sei}}{dt}}={-A}_{sei}·c_{sei}·exp\left[-{\displaystyle \frac{E_{a,sei}}{RT}}\right]$	$Q_{sei}={-H}_{sei}{·W}_{sei}·R_{sei}$
negative-electrolyte reaction	100–130 °C	$$\begin{gathered} R_{ne}={\displaystyle \frac{d{t}_{sei}}{dt}}=-{\displaystyle \frac{d{c}_{ne}}{dt}} \\ ={-A}_{ne}·c_{ne}·exp\left[-{\displaystyle \frac{t_{sei}}{t_{sei,ref}}}\right] \\ ·exp\left[-{\displaystyle \frac{E_{a,ne}}{RT}}\right] \end{gathered}$$	$Q_{ne}=-H_{ne}{·W}_{ne}·R_{ne}$
positive-electrolyte reaction	150–300 °C	$R_{pe}={\displaystyle \frac{d{c}_{pe}}{dt}}=A_{pe}·\alpha ·\left(1-\alpha \right)·exp\left[-{\displaystyle \frac{E_{a,pe}}{RT}}\right]$	$Q_{pe}=H_{pe}{·W}_{pe}·R_{pe}$
electrolyte decomposition	Around 200 °C	$R_e={\displaystyle \frac{d{c}_e}{dt}}=-A_e·c_e·exp\left[-{\displaystyle \frac{E_{a,e}}{RT}}\right]$	$Q_e={-H}_e{·W}_e·R_e$
ISC	Around 130 °C	$$\begin{gathered} R_{ISC}=-{\displaystyle \frac{dSOC}{dt}} \\ ={ISC}_{cond}{·SOC·A}_{ISC}·exp\left[-{\displaystyle \frac{E_{a,ISC}}{RT}}\right] \end{gathered}$$	$$\begin{gathered} Q_{ISC}=H_{ISC}·R_{ISC} \\ =V·C·3600·\eta {·R}_{ISC} \end{gathered}$$

. The physical and kinetic parameters used in the heat generation equations are presented in

Table 3. Parameters for the heat generation equations of TR reactions [22,46,47].

Parameters	Description	Unit
$A_{sei/ne/pe/e/ISC}$	Frequency factor of the reaction	$s^{-1}$
$E_{a,sei/ne/pe/e/ISC}$	Activation energy of the reaction	J/mol
$c_{sei/ne/e}$	Initial dimensionless content	-
$\alpha$	Conversion degree of the positive electrode material	-
$t_{sei,ref}$	Initial dimensionless SEI layer thickness	-
$H_{sei/ne/pe/e/ISC}$	Heat of reaction	J/kg
$W_{sei/ne/pe/e}$	Material content	kg/m3
$V$	Battery nominal voltage	V
$C$	Battery capacity	Ah
$\eta$	Efficiency factor	-
${ISC}_{cond}$	Parameter to control heat generation by ISC: $$\begin{gathered} - {ISC}_{cond}=0, T<130 °\mathrm{C} \\ - {ISC}_{cond}=1, T\ge 130 °\mathrm{C} \end{gathered}$$	-

.

The immense heat generation during TR poses significant safety risks in LIBs. As the battery undergoes TR, exothermic reactions lead to rapid and uncontrollable temperature increases, often exceeding 1000 °C in the absence of cooling strategies [22]. This excessive heat can result in the rupture of the battery casing, release of toxic gases, fires, and explosions. Such catastrophic events can be triggered by relatively minor incidents, such as overcharging or physical damage, highlighting the critical need for preventive measures. Advanced BTMS are essential to mitigate these risks by controlling the temperature within safe operating limits, preventing TR initiation and suppressing TRP. By incorporating advanced cooling strategies such as liquid cooling, PCM cooling, and hybrid cooling, BTMS can prevent TR and significantly reduce the chances of a domino effect in multi-cell configurations, enhancing the safety and reliability of battery-powered systems. Therefore, reviewing robust thermal management solutions is paramount to ensuring the safe operation of LIBs, particularly in high-energy applications like EVs.

2.4. Comparison Between Traditional and Advanced Cooling Strategies

Traditional BTMS primarily utilize air cooling, owing to their simplicity and low cost [48,49]. Air cooling relies on natural or forced convection to dissipate heat from battery modules [50,51]. However, the relatively low thermal conductivity and heat capacity of air limits its cooling effectiveness, especially in high-energy battery systems [52,53,54]. Consequently, air cooling often results in non-uniform temperature distributions and insufficient heat removal under abusive conditions, increasing the risk of TR.

In contrast, advanced cooling strategies, such as liquid cooling and PCM cooling, offer significantly higher heat transfer rates and more precise temperature control. These methods improve temperature uniformity, rapidly absorb and dissipate heat, and can suppress or delay TR and its propagation [55,56,57]. The limitations of traditional air cooling thus motivate the exploration and development of these advanced thermal management solutions, which are the focus of the subsequent sections.

3. Advanced Cooling Strategies for TR Prevention and Suppression in LIBs

This section presents a comprehensive review of advanced BTMS designed to prevent and suppress TR and TRP in LIB modules. With the increasing energy density and power demand in modern battery systems, traditional cooling methods are often insufficient to address the safety challenges posed by rapid heat generation and propagation. Therefore, advanced cooling strategies have emerged as critical components in enhancing thermal stability and operational safety. These advanced cooling strategies include liquid cooling methods, such as indirect liquid cooling, water mist cooling, and immersion cooling, which are discussed in Section 3.1. PCM cooling systems are detailed in Section 3.2. Finally, hybrid cooling approaches combining multiple mechanisms to maximize heat dissipation efficiency are presented in Section 3.3.

3.1. Liquid Cooling

3.1.1. Indirect Liquid Cooling

Indirect liquid cooling is a prevalent thermal management technique extensively studied for LIB modules, primarily due to its superior heat dissipation capabilities compared to traditional air cooling [58]. Numerous studies have investigated its potential for suppressing TRP events. Fu et al. demonstrated that using mini-channel cold plates based on indirect liquid cooling effectively delayed and even completely suppressed TRP, depending on the coolant velocity and channel width, as shown in Figure 2. Specifically, the results show that, by applying cold plates with mini-channels, the cooling system effectively delayed the occurrence of TRP at low coolant flow velocities of 0.2 m/s. At this flow rate, only battery 1 and battery 2 underwent TR, with maximum temperatures of approximately 360 °C and 200 °C, respectively, while battery 3 remained below the critical temperature threshold, reaching a maximum of 88 °C without experiencing TR. When the coolant velocity was further increased to 0.3 m/s, TR was confined solely to battery 1, the temperature of battery 2 remained below 88 °C, and battery 3 maintained a temperature below 55 °C, indicating complete suppression of TRP at this flow velocity. In addition, the mini-channel width significantly affected the cooling performance. At a fixed coolant flow velocity of 0.2 m/s, TRP was eliminated when the mini-channel width was $\ge$7 mm [59].

Similarly, Ke et al. highlighted that high coolant flow rates through serpentine-channel cooling significantly mitigated TRP, as presented in Figure 3. The results demonstrate that coolant flow rates of 96 L/h or higher effectively prevented TRP, maintaining the temperature of the triggered TR cell at around 450 °C while keeping neighboring cells below 80 °C. In contrast, lower flow rates of 0 L/h, 32 L/h, and 64 L/h were insufficient to suppress TRP, allowing for propagation to the adjacent cells. The study also showed that the expulsion of high-temperature electrolytes from the battery’s positive side when TR is activated plays an important role in propagating TR to adjacent cells. This electrolyte expulsion mechanism was identified as the main contributor to TRP, while conduction and radiation played only minor roles [60].

However, despite their promising performance in TR prevention and TRP suppression, indirect liquid cooling systems present inherent challenges. Rui et al. pointed out that, while indirect liquid cooling reduced peak heat flux relatively, it was insufficient to fully prevent TRP without complementary thermal insulation. Aerogel insulation layers significantly enhanced suppression effectiveness, underlining the limitation of relying solely on indirect liquid cooling. Specifically, compared to the basic configuration, a configuration with indirect liquid cooling reduced the peak heat flux of a TR cell and adjacent cells from 885.7 W to 848.2 W. However, this decrease was not sufficient to prevent TRP. When a 1 mm thick aerogel insulation layer was introduced between cells in the battery module, the TRP was significantly slowed, with the average interval between TR events extending from 84.4 s to 384.2 s. Ultimately, the combined application of liquid cooling and thermal insulation layers successfully prevented TRP, maintaining the temperature of the triggered TR cell at around 500 °C while keeping adjacent cells below 80 °C [61].

Similarly, Liu et al. conducted a comprehensive study on the suppression of TRP in LIB packs using mini-channel cold plates and insulation layers. Simulations showed that adding insulation layers between battery cells significantly delayed the onset of TR in adjacent cells. Specifically, when 2 mm and 3 mm aerogel insulation layers were introduced, the time for TR to trigger in the adjacent cell was delayed by 312 s and 944 s, respectively. In addition, the TRP to the entire battery pack was delayed by 1293 s, which far exceeded safety standards. Furthermore, the proposed cooling strategy controlled the average temperature of adjacent cells not to exceed 80 °C, the LIB pack is protected and the propagation of TR is completely prevented [62]. These results emphasize the role of combining insulation layers with indirect liquid cooling systems for effective TR prevention and TRP suppression in LIB modules.

In addition, Yang et al. noted that, although indirect liquid cooling based on cold plates relatively effectively managed temperatures and suppressed TRP, coolant configurations could unintentionally transfer heat and trigger adjacent cell failures due to direct cell-to-cell conduction facilitated by coolant channels. Specifically, the study evaluated the TRP inhibiting effectiveness of BTMS using indirect liquid cooling with cold plates placed at various locations inside the battery pack, including bottom (BTMS-1a), side (BTMS-1b), and inter-cell (BTMS-1c) cold plates, as depicted in Figure 4. The research results show that BTMS-1a demonstrated the highest cooling rate, reaching 2.03 kW. Thus, maintaining the best temperature uniformity within the battery pack with 13.4 °C at a 5C discharge rate. However, BTMS-1a had the shortest propagation time of 74.4 s due to the direct contact between cells and the high-temperature coolant transferring heat to other cells, leading to TR triggering. When the coolant flow rate in each cold plate was increased to 0.75 L/min, BTMS-1b could further delay the occurrence of TR to 82.6 s. Notably, at a similar flow rate, BTMS-1c significantly enhanced cell cooling, completely inhibiting TRP. In this case, the maximum temperature of the triggered TR cell reached around 1000 °C, while the adjacent second cell remained near 100 °C and other neighboring cells were maintained at approximately 60 °C [63]. These research results demonstrate that optimization of cold plate layout and coolant flow plays an important role in TR prevention and TRP suppression in the LIB modules of indirect liquid cooling systems.

Recent advancements aim to address these shortcomings through novel structural designs. For example, Du et al. presented a novel BTMS that incorporates a triply periodic minimal surface (TPMS) porous structure in liquid cooling channels to enhance heat dissipation and reduce the risk of TR, according to Figure 5. The results show that the battery pack using the primitive porous liquid cooling structure showed a 12.43% reduction in maximum temperature and an 8.41% reduction in temperature difference compared to the straight-tube channel. In addition, under extreme operating conditions up to 50 °C, the primitive porous structure still maintained the maximum temperature of the battery pack below 54.66 °C, and the temperature difference did not exceed 3.76 °C, effectively preventing TR events [64].

Critically evaluating these findings reveals significant knowledge gaps regarding the long-term reliability of indirect liquid cooling, particularly concerning coolant leakage, corrosion, and material compatibility under prolonged operation and thermal abuse scenarios. The complexity and increased parasitic energy demands associated with pumping coolant through intricate channels also raise concerns regarding practical feasibility and economic viability in commercial EV battery applications. Addressing these gaps will require future research to optimize coolant materials, refine cooling channel architectures, integrate advanced insulation materials, and comprehensively test system performance under realistic operational conditions.

3.1.2. Water Mist

Water mist cooling has gained attention as a promising active cooling strategy for TR prevention and TRP suppression in LIB modules, attributed primarily to its high latent heat of vaporization and rapid cooling capacity. By using pure water with non-conductive properties, water mist cooling ensures the safety and reliability of LIBs cooling during TR events. Experimental studies have demonstrated notable cooling efficacy of water mist cooling under various TR conditions. Ma et al. investigated the cooling effect of water mist to suppress TRP in LIB modules under various conditions, as demonstrated in Figure 6. The study found that water mist applied effectively reduced the temperature rise in the cells during the TR process. When water mist was released 30 s before the onset of TR, it successfully suppressed the occurrence of TR. However, when the rate of temperature increase reached 2 °C/s, water mist could not completely suppress TR, but significantly reduced the temperature increase and surface temperature of the cells [65].

Liu et al. conducted an experimental study on the cooling effect of fine water mist (FWM) to mitigate TR in LIBs. The results demonstrate that FWM significantly enhanced the cooling capacity, achieving a maximum improvement of 29.5% compared to air cooling. When the FWM flow rate increased from 0.1 mL/s to 0.8 mL/s, the cooling time to reduce the maximum temperature of the battery during TR to below 100 °C was reduced by 94.5 s, from 496.0 s to 401.5 s [66].

Despite these promising results, water mist cooling faces critical limitations, notably in densely packed battery modules, where the efficacy of water mist penetration is significantly compromised by flame-induced buoyancy and gas ejection phenomena. These limitations suggest that conventional water mist delivery systems might not provide sufficient penetration depth or uniform cooling under severe TR scenarios. Therefore, Liu et al. evaluated and improved the efficiency of water mist cooling by determining a critical water consumption threshold necessary for successful TRP suppression. The results demonstrate that water mist provided excellent cooling performance, achieving cooling rates exceeding 100 °C∙s⁻¹ and reducing the cell surface temperature to approximately 100 °C within a few seconds. At the same time, the research results also show that a critical water consumption rate of 1.95 × 10⁻⁴ kg∙Wh⁻¹ was sufficient to suppress TRP, even under extreme conditions where four out of five cells experienced TR [67]. The above research results show that identifying a critical water consumption threshold is effective for successful TRP suppression; however, this indicates that effective water mist cooling systems could demand substantial volumes of water, consequently increasing system complexity, cost, and weight.

Furthermore, recent innovations have focused on enhancing water mist cooling through intermittent spraying and hybrid approaches. Yao et al. demonstrated superior performance using intermittent mist spray cycles combined with extinguishing agents like C₆F₁₂O, markedly improving fire suppression effectiveness and significantly controlling battery temperature increases. This intermittent spraying approach not only enhanced cooling efficiency, but also optimized water consumption. Specifically, the study showed that integrating C₆F₁₂O with water mist significantly enhanced the fire suppression ability compared to continuous water spray alone, effectively extinguishing the battery flame within 1 s. Notably, the intermittent water mist spray mode with a duty cycle (DC) of 0.5 and a cycle period (Pt) of 2 s showed superior cooling performance, limiting the maximum battery temperature to 115.9 °C and demonstrating the highest heat suppression ability. In addition, a longer intermittent cycle (DC = 0.1, Pt = 20 s) provided a prolonged cooling time, effectively reducing the battery temperature to below 50 °C, completely preventing the spread of uncontrolled TR [68].

Similarly, Xu et al. investigated the cooling effect of water mist with additives to suppress TR events in LIBs. The results show that water mist combined with additives such as Tween 20 (Polyoxyethylene (20) sorbitan monolaurate—C₁₈H₃₄O₆(C₂H₄O)₂₀) or 1-heptanol—significantly improved the cooling efficiency during TR events. In particular, water mist with Tween 20 provided the highest cooling efficiency, dissipating 12.46 ± 0.20 kJ of heat, an increase of 23.6% compared to pure water. This improvement was reflected in a reduction in temperature, with the surface temperature of the battery decreasing by 8.33 ± 0.28 °C/s. Furthermore, when water mist is applied, the cooling effect significantly reduces the fire extinguishing time, reducing the burn time from 92 s in the test without water mist to just 5 s [69].

Critically assessing the available studies, there remain significant gaps in understanding the optimal mist particle size, delivery frequency, and distribution mechanisms within battery packs. Additionally, water mist cooling systems must ensure robust electrical insulation to prevent short circuits or leakage currents that could compromise safety. Future research should prioritize developing tailored mist delivery systems that ensure deep penetration and uniform coverage, as well as investigating integrated cooling strategies that combine water mist cooling with other reliable extinguishing agents. Addressing these gaps is crucial for enhancing the practical feasibility, effectiveness, and scalability of water mist cooling systems in commercial LIB applications.

3.1.3. Immersion Cooling

Immersion cooling, also known as direct liquid cooling, offers an effective thermal management solution by submerging battery cells in dielectric fluids to ensure rapid and uniform heat dissipation. This approach eliminates thermal interface resistance and facilitates high-efficiency heat removal, making it particularly effective for managing extreme thermal events such as TR and TRP. Ye et al. conducted an experimental investigation on immersion cooling to inhibit the TR behavior of LIBs and evaluate the suppression of TRP in battery modules. The findings indicate that a full 70 mm immersion significantly delayed the onset of TR, extending the safety valve rupture-to-TR trigger interval from 70 s (at 20 mm immersion) to 312 s and reducing the peak TR temperature to 306 °C. Additionally, cells in 100%, 75%, and 50% SOC designate greater battery capacity, which elevates the risk of TR and thermal hazards. Furthermore, immersion cooling effectively inhibited the inter-cell TRP with peak temperature reductions of adjacent cells in 100%, 75%, and 50% SOC modules of 118.2 °C, 121.4 °C, and 111.5 °C, respectively [24].

In addition, several studies emphasize the role of fluid selection in optimizing the cooling efficacy of immersion cooling. Liu et al. investigated the use of immersion cooling with fluorinated liquids (HFE-7000 and SF33) to mitigate uncontrolled TRP in LIB packs, as shown in Figure 7. The results show that immersion cooling provided outstanding heat transfer rates and significantly reduced the surface temperature of the TR cell, with temperatures maintained at 518.8 °C and 448 °C for HFE-7000 and SF33, respectively. Furthermore, the immersion cooling system completely inhibited the propagation of TR, with the temperature maintained for neighboring cells at around 34 °C [25].

Similarly, Sadar et al. investigated the use of Transformer Oil (TO) immersion cooling as a method to prevent TR in LIB packs under extreme conditions, as depicted in Figure 8. The results from TR testing show that the center cell temperature in the TO immersion setup reached a maximum of 222 °C, well below the critical temperature for TR, which typically exceeds 300 °C. Notably, TO prevented TR after 18,000 s, providing significant time for safe intervention. This extended time allows for the more effective identification of potential problems before catastrophic TR occurs [70].

Furthermore, some studies have further explored the flow configuration in immersion cooling setups, aiming to optimize the distribution of the fluid to achieve better temperature uniformity and lower maximum temperatures in LIB modules. Yang et al. examined the effectiveness of single-phase immersion cooling (SPIC) in managing the thermal behavior of 280 Ah LiFePO₄ batteries, focusing on the effects of different flow arrangements, including jet impingement, same sides and opposite sides, as demonstrated in Figure 9. The results show that jet impingement cooling provided the best temperature control, with a maximum temperature reduction of 1.29 °C compared to the same side arrangement. In addition, the temperature difference was 1.32 °C lower in the jet impingement system. The study also highlighted that no ignition or explosion occurred during the temperature spike during immersion cooling, with the maximum temperature during the temperature spike reaching 241.9 °C, demonstrating the safety and reliability of SPIC in preventing catastrophic TR events in large-scale battery systems [71].

Similarly, Patil et al. examined the use of dielectric fluid immersion cooling for LIBs, focusing on optimizing the flow configuration and tab cooling to improve cooling efficiency and prevent TR, as illustrated in Figure 10. The study found that LIB cells immersed in a dielectric fluid flowing from bottom to top assisted with tab cooling demonstrated a 46.8% reduction in maximum temperature at the positive tab at a 3C discharge rate compared to natural convection. This method resulted in a significant improvement in cooling performance, keeping the battery temperature below 40 °C even at a high 5C discharge rate. Furthermore, when simulating thermal abuse conditions such as ISC, the peak temperature of the battery pack was limited to 341.7 °C, successfully preventing the propagation of TR throughout the cells [72]. These insights underscore the importance of optimized fluid dynamics in maximizing the effectiveness of immersion cooling systems.

While immersion cooling systems demonstrate exceptional performance in heat mitigation, several practical and safety challenges must be addressed. The high cost and limited environmental compatibility of many dielectric fluids, such as fluorinated liquids, restrict their widespread adoption. Chemical stability, potential toxicity, and disposal issues raise additional environmental and regulatory concerns. Moreover, ensuring structural sealing against leaks is vital, especially in two-phase systems where pressure fluctuations occur due to phase change. A critical safety consideration involves the dielectric strength and material compatibility of immersion fluids with cell packaging and electrical components. Any breach in insulation could compromise safety and lead to short circuits, particularly in systems exposed to thermal abuse. Therefore, the dielectric properties of fluids must be rigorously validated under real-world conditions, including long-duration cycling and extreme temperatures. Although current research demonstrates promising results, further studies are needed to evaluate the long-term reliability of immersion-cooled LIB packs, the environmental impact of different coolant formulations, and the trade-offs between performance, safety, and cost. Future directions should focus on developing biodegradable or low-toxicity dielectric fluids, enhancing sealing techniques, and exploring hybrid cooling strategies that incorporate immersion cooling with supplemental insulation or PCM for comprehensive TR prevention and TRP suppression.

3.2. PCM Cooling

PCM cooling has emerged as a promising passive thermal management strategy due to its ability to absorb large amounts of heat during phase transitions, thereby delaying or suppressing the rapid temperature escalation associated with TR and TRP in LIBs. Unlike active systems, PCM cooling does not require external power input, making it particularly attractive for compact energy-efficient applications. Liu et al. conducted numerical simulations to investigate the effectiveness of paraffin phase change material (PPCM) as a thermal safety measure against uncontrolled TRP in LIB packs, as shown in Figure 11. The results show a clear correlation between the thickness of PPCM and its ability to slow down TRP. Specifically, the TR onset in the adjacent cell was significantly delayed to 64 s, 266 s, and 414 s, corresponding to PPCM thickness increases of 1.8 mm, 3.6 mm, and 5.4 mm, respectively. Furthermore, combining PPCM with insulation layers provided additional protection and extended the TR onset delay in the adjacent cell to 798 s, corresponding to an approximately 43-fold improvement over the unprotected LIB pack [25].

Yang et al. evaluated the use of PCM cooling in BTMS to mitigate uncontrolled TRP, as presented in Figure 12. The results show that PCM cooling effectively absorbed and managed heat in TRP, reducing the maximum temperature to 913.5 °C, which was significantly lower than the temperature achieved with only using insulation materials of 1165.9 °C. In addition, PCM cooling could delay the propagation time of TR until 80.8 s due to the advantages of large latent heat and a high heat absorption rate [63].

Sadar et al. explored the use of PCM cooling to prevent TR in LIB packs under extreme conditions, as demonstrated in Figure 13. During TR testing, the center cell temperature in the PCM cooling setup peaked at 434 °C, significantly lower than the 560 °C observed in the natural convection setup. Additionally, PCM cooling significantly delayed the time to the TR event by 1400 s, 500 s longer than the natural convection battery pack configuration [70].

Zhi et al. studied the application of a hydrated salt-based composite PCM consisting of a mixture of Na₂SO₄⋅10H₂O and KAl(SO₄)₂⋅12H₂O to the thermal management of LIBs and to mitigate TR, as shown in Figure 14. The results show that the maximum battery temperature decreased by 20.4%, from 65.95 °C, with air cooling to 52.47 °C with PCM cooling, and the temperature difference decreased by 59.7%, from 5.46 °C to 2.20 °C, respectively, during a 0.5C charge and a 1.5C discharge cycle. The PCM cooling also demonstrated its effectiveness in preventing TRP, with the peak temperature of the first cell during TR reduced from 427 °C (without PCM) to 383 °C (with PCM). Furthermore, the temperatures of the adjacent cells with the TR triggering cell were significantly reduced by the PCM barrier, completely preventing TR events [73]. These results demonstrate the feasibility of tailoring PCM compositions to enhance both thermal regulation and TRP suppression.

Despite these advantages, PCM systems are constrained by inherently low thermal conductivity, which limits the rate of heat absorption and dissipation. This drawback becomes especially pronounced under high-power or prolonged abuse conditions where heat generation exceeds the phase change capacity of the PCM. While additives such as expanded graphite, metal foams, or thermally conductive fillers have been employed to enhance thermal conductivity, their integration must be balanced against material stability, cost, and added weight. In addition to performance concerns, safety and material compatibility issues must be considered. Paraffin-based PCMs, although being commonly used, may pose fire risks if containment is breached. Hydrated salts, while non-flammable, can exhibit phase separation or subcooling, which reduces reliability over long-term cycles. Therefore, selecting PCMs with stable thermal cycling properties, high latent heat, and minimal volumetric expansion is crucial for system durability and safety. Critically, PCM systems alone cannot maintain thermal equilibrium during extended thermal events without supplemental dissipation. As such, PCM cooling is best suited for transient heat suppression or as a component in hybrid systems. Future research should aim to optimize PCM formulations for enhanced conductivity and cycling stability, develop composite materials with integrated structural and thermal functions, and validate long-term performance under realistic operational scenarios. Additionally, standardization in testing methodologies for PCM-enhanced BTMS will be essential to accelerate their adoption in practical battery applications. Furthermore, research into combining PCM cooling with advanced cooling strategies such as liquid cooling will further improve the goal of temperature control and prevention of TR and TRP events in LIB modules.

3.3. Hybrid Cooling

Hybrid cooling systems, which integrate both active and passive cooling mechanisms, have gained prominence as highly effective strategies for suppressing TR and TRP in LIB modules. By combining the high latent heat absorption of PCM with the continuous heat removal capabilities of liquid cooling, hybrid systems aim to achieve optimal thermal regulation, temperature uniformity, and safety performance under both normal and abuse conditions. Xiao et al. investigated a hybrid cooling strategy that combined composite phase change material (CPCM) and liquid cooling to enhance the thermal safety of LIB packs under TRP conditions, as depicted in Figure 15. The research results show that increasing the coolant flow rate significantly delayed TRP, with the propagation time extended from 17.8 s at 0.009 m/s to 32.9 s at 0.029 m/s. Additionally, increasing battery spacing delayed propagation and altered its path, while decreasing the thermal conductivity of the CPCM substantially reduced TRP due to limited heat transfer. Conversely, increasing the number of cooling channels had a limited effect on TRP time, with the total TRP time changing from 25.5 s to 27 s when the number of cooling channels increased between 1 and 5 [46].

Luo et al. proposed an adaptive hybrid cooling strategy to mitigate TR in LIB modules by integrating liquid cooling and PCM cooling with considerations of buoyancy-driven natural convection in PCM, as illustrated in Figure 16. The findings indicate that the proposed hybrid cooling approach extended the TRP interval by 60.6% (85.9 s) and reduced total coolant flow by 31.7%. In addition, the implementation of hybrid cooling improved the cooling performance without causing additional parasitic energy consumption or compromising the energy density of the battery pack [29].

Lu et al. proposed a thermal management system using a hybrid cooling strategy combining PCM cooling based on a paraffin/Al-foam composite and liquid microchannels to suppress TR in LIB packs, as depicted in Figure 17. The research results demonstrate that the proposed hybrid cooling system effectively suppressed TR, with the temperature of neighboring cells only increasing to about 80 °C in the first 600 s, which could not trigger a TR at a liquid flow rate of 0.01 L/min. The proposed hybrid cooling system outperformed individual PCM or liquid microchannel cooling methods and significantly improved battery safety by suppressing TR across multiple cells [74]. These findings underscore the hybrid approach’s ability to balance energy efficiency, safety, and thermal performance.

Beyond thermal control, advanced hybrid systems have been designed with multifunctional components to improve system compactness and heat spreading. Ouyang et al. proposed a novel hybrid cooling system that integrates CPCM, aerogel for thermal insulation, and a cooling plate with nanofluid to enhance both heat dissipation and insulation, as demonstrated in Figure 18. The research showed that the proposed hybrid cooling system efficiently mitigates TRP and reduces the maximum battery temperature from 740.35 °C to 55.19 °C under normal heat dissipation conditions. Under extreme conditions, the maximum battery temperature was reduced from 707.36 °C to 107.19 °C. In addition, the cooling performance of the hybrid cooling system was further optimized with the best combination of factors, including a flow rate of nanofluid of 0.05 m/s, nanoparticle volume fraction of 1%, and a thickness of the PCM and insulating material of 3 mm, which resulted in a 23% reduction in maximum battery temperature and a 22% reduction in energy consumption compared to the baseline configuration [75].

Similarly, Zhang et al. proposed a hybrid cooling strategy combining PCM and liquid cooling to mitigate uncontrolled TRP in LIB modules, as shown in Figure 19. The research results demonstrate that the proposed hybrid cooling system effectively controlled the battery temperature, with the maximum temperature during a 5C discharge cycle maintained at 39.8 °C, compared to 51.7 °C for the only liquid-cooled system. The hybrid cooling system also slowed down the uncontrolled TRP with a delayed time of 460 s between the two cells, while the liquid-cooled system had nearly simultaneous uncontrolled TR events. In addition, the higher PCM thermal conductivity (0.6 W/m·K) improved the conventional cooling performance, but increased the heat transfer rate, resulting in uncontrolled TRP in 5.6 min. Furthermore, increasing the water velocity in the cooling channel to 0.01 m/s successfully prevented uncontrolled TRP in the battery module [76].

Wang et al. studied a hybrid cooling strategy that combines PCM cooling with wavy microchannel cooling plate (PCM-WMCP) to enhance cooling efficiency and temperature uniformity in LIBs, as presented in Figure 20. The numerical study found that the hybrid cooling system significantly reduced the maximum battery temperature compared to the air-cooled (AC) system and the PCM-only system. Specifically, at a discharge rate of 1.0C, the maximum battery temperature decreased from 59.49 °C (AC) to 43.65 °C (PCM-WMCP), and, at a discharge rate of 3.0C, the battery temperature decreased from 79.66 °C (AC) to 49.45 °C (PCM-WMCP). Additionally, when 10% graphite was added to the PCM to form a composite PCM (CPCM(B)), the system achieved a more uniform temperature distribution, reducing the temperature difference between the maximum and minimum battery temperatures from 2.14 °C to 1.61 °C. Furthermore, the hybrid CPCM(B)-WMCP system also demonstrated superior performance in preventing TR, with the time to reach the critical temperature for TR being 650 s slower than the PCM-only system [77].

Gu et al. proposed a novel hybrid cooling that integrates CPCM (paraffin wax and boron nitride) and a variable wall liquid cooling plate (LCP), as depicted in Figure 21. The results show that the implementation of the variable wall LCP resulted in a significant reduction in the maximum battery temperature, which was reduced by 1.81 °C compared to the standard LCP. Furthermore, the study found that maintaining a flow rate above 0.89 g/s was sufficient to keep the maximum battery temperature below 40 °C during both the charge and discharge cycles, providing a practical hybrid cooling solution to prevent TR in high-performance batteries [78].

Zonouzi et al. explored an innovative hybrid cooling approach to the thermal management of LIBs by integrating helical tube liquid cooling with PCM cooling. The results demonstrate that the hybrid cooling system outperformed individual cooling methods, with the maximum battery temperature after 750 s being reduced to 34.45 °C in the hybrid cooling system, compared to 38.55 °C with only PCM cooling and 36.35 °C with only liquid cooling. The study also highlighted the positive impact of reducing the helical tube pitch, which significantly improved the cooling performance by providing a larger contact area for heat exchange and contributing to the effective prevention of potential TR events in LIB application systems [79].

Liu et al. evaluated a hybrid cooling system combining a mini-channel cold plate (MCP) with PCM cooling for LIBs, especially under a high discharge rate and TR conditions. The results show that, during a high discharge rate of 2.5C, the optimized hybrid cooling structure maintained the battery temperature at 44.37 °C and a temperature difference between cells of no more than 2.01 °C, even at low coolant flow rates. Under external short-circuit conditions, the optimized structure maintained the temperature of the battery pack at 43.22 °C, providing effective thermal management. Most notably, when TR occurs due to an ISC, the optimized hybrid cooling configuration can delay TR propagation by up to 521 s, significantly improving safety by preventing TR from spreading to adjacent cells [80].

Ji et al. developed a novel hybrid cooling strategy combining phase change composite materials (PCCM) with a liquid cooling plate to suppress TRP in high-specific-energy ternary LIBs. The experimental results demonstrate that when a 2 mm thick PCCM was integrated with a liquid cooling plate, the TR in a 60 Ah battery module was successfully mitigated. Specifically, during the TR event, the TR trigger cell reached a temperature of approximately 650 °C; however, due to the heat absorption capacity of the PCCM and the cooling effect from the liquid cooling plate, the temperature of the adjacent cell remained under control and remaining below 100 °C throughout the process, effectively suppressing TRP [81].

Despite these advancements, hybrid cooling systems present several engineering and implementation challenges. The integration of multiple components increases system volume, design complexity, and manufacturing cost. Inefficient thermal coupling between PCM and liquid channels can limit their effectiveness during rapid thermal transients, while the saturation of PCM over extended events may diminish passive cooling capacity. Furthermore, additional structural materials may reduce energy density, a critical consideration in space-constrained EV applications. From a safety perspective, hybrid systems must be carefully designed to avoid coolant leakage, electrical shorting, and material degradation under thermal stress. Dielectric compatibility of coolants, structural sealing techniques, and insulation integrity are essential for maintaining electrical isolation and mechanical reliability over time. To address these issues, future research should prioritize the co-optimization of hybrid system geometries, including cooling channel layouts and PCM distributions to maximize thermal synergy and spatial efficiency. Investigating advanced composite PCMs with self-healing or thermally triggered reusability may provide further breakthroughs. Finally, comprehensive long-term performance validation under real-world duty cycles will be critical for enabling the safe and scalable deployment of hybrid BTMS in high-energy LIB applications.

4. Summary and Recommendations

TR events remain one of the most critical safety concerns in LIB applications. TR is often triggered by mechanical, electrical, or thermal abuse and is characterized by rapid heat generation, gas release, and potential fires or explosions. As LIBs are increasingly deployed in high-energy and high-power applications, the demand for advanced BTMS capable of preventing and suppressing TR events has become paramount. Advanced BTMS technologies offer promising solutions by enhancing heat dissipation, maintaining battery temperature uniformity, and responding flexibly to thermal anomalies. In this context, the present review focuses on the recent studies over the past 5 years on advanced BTMS for the prevention and suppression of TR events and TRP in LIB modules. The advanced cooling strategies summarized and evaluated in the present review include liquid cooling methods such as indirect liquid cooling, water mist cooling, and immersion cooling, followed by PCM cooling and, finally, hybrid cooling systems. The following summarizes the main conclusions and recommendations of the current review:

(a)

Indirect liquid cooling is a widely adopted thermal management technique that offers superior heat dissipation compared to traditional air-cooled methods, primarily due to the higher thermal conductivity and specific heat capacity of liquid coolants such as water or water–glycol mixtures. This approach enables more effective temperature regulation and improved thermal uniformity across the battery module. However, the presence of multiple thermal interfaces between the battery cells and cold plates introduces significant thermal resistance, limiting the cooling effectiveness during critical (TR) events. Additionally, indirect liquid cooling systems often involve complex structural designs, increasing the risks related to coolant leakage, corrosion, and higher parasitic energy consumption. These factors contribute to increased system weight, maintenance demands, and overall cost, posing challenges for scaling to next-generation high-power LIB applications. Therefore, while indirect liquid cooling remains effective under standard operating conditions, its application for TR prevention and suppression in future high-capacity systems requires further optimization in terms of system design, coolant management, and safety validation.

(b)

Water mist cooling has emerged as a promising strategy to mitigate TR and its propagation in LIBs due to its high latent heat of vaporization and rapid heat absorption through droplet evaporation. By cooling both the ambient environment and the battery surface, water mist can significantly reduce the temperature rise during TR events. Furthermore, water mist systems offer environmental advantages, being non-toxic, readily available and cost-effective. However, their effectiveness can be compromised in dense battery modules where flame-induced buoyancy and gas ejection limit mist penetration to inner cells. Additionally, large water volumes are often required to sustain sufficient cooling performance, which increases system complexity and weight. Despite these limitations, water mist cooling demonstrates strong potential for TR suppression, and further improvements in delivery methods, droplet sizing, hybrid agent use, and real-time thermal control are needed to enable its broader application in advanced battery thermal management systems.

(c)

Immersion cooling, also known as direct liquid cooling, offers a highly effective approach for preventing and suppressing TR and TRP in high-energy LIB systems. By directly submerging battery cells in dielectric fluids, immersion cooling eliminates thermal interface resistance and achieves rapid, uniform heat dissipation. Single-phase immersion cooling provides flexible system design options, while two-phase immersion cooling leverages latent heat for enhanced cooling performance. However, challenges such as the high cost, environmental sensitivity, and chemical stability requirements of dielectric fluids, along with concerns over leakage, material compatibility, and system complexity restricted widespread adoption. Additionally, ensuring robust structural sealing is critical, particularly in two-phase systems where boiling-induced pressure variations occur. Despite these issues, innovative designs such as tab-assisted cooling, optimized coolant flows, and improved immersion architectures show strong potential to make immersion cooling a reliable and scalable solution for enhancing LIB safety against TR and TRP events.

(d)

PCM cooling presents a promising passive thermal management strategy for LIBs, particularly in compact and safety-critical applications. PCMs absorb substantial amounts of heat through latent heat during phase transitions, effectively delaying or suppressing rapid temperature rises without external energy input. Paraffin- and hydrated salt-based PCMs have demonstrated significant benefits in retarding TRP by maintaining temperature uniformity within battery packs. Enhancements through composite formulations, such as adding expanded graphite or metallic fillers, further improve thermal conductivity and structural stability. However, the inherently low thermal conductivity of pure PCMs limits their standalone performance, especially under severe abuse conditions. Moreover, PCM systems lack active heat dissipation mechanisms, making them more suitable for low to moderate power densities or as supplementary layers in hybrid cooling architectures. Future developments in high-conductivity composite PCMs and integration with active cooling systems are critical to overcoming current limitations and unlocking the full potential of PCM cooling for advanced battery thermal management.

(e)

Hybrid cooling systems that integrate passive and active thermal management techniques have emerged as highly effective solutions for mitigating TR and TRP risks in LIB modules. By combining the latent heat absorption of PCMs with the continuous heat dissipation of liquid cooling, hybrid systems achieve superior control over both peak temperatures and temperature uniformity. Studies have demonstrated that hybrid configurations can significantly delay TRP events and prevent adjacent cells from reaching critical failure conditions. However, the complexity of integrating active and passive components increases design difficulty, system volume, and cost. Inefficient thermal coupling between PCM layers and liquid cooling structures can limit overall system performance, particularly during rapid TR events. Additionally, PCM saturation over prolonged abuse conditions may reduce cooling effectiveness. Therefore, future research must focus on optimizing composite materials, refining liquid channel geometries, and developing modular, lightweight, hybrid architectures to fully realize the potential of hybrid cooling in next-generation BTMS designs.

(f)

In addition, thermal insulation layers have shown substantial potential as a passive mitigation strategy for TRP in LIB systems. By introducing low thermal conductivity materials such as aerogels or glass fiber composites between adjacent cells, insulation layers act as thermal barriers, effectively delaying or even preventing the spread of heat generated during a TR event. Moreover, when combined with active cooling systems such as cold plates or immersion cooling setups, insulation layers allow for the cooling system to have more time to remove heat before it reaches critical levels. Despite their advantages, insulation layers can accumulate heat over prolonged operation and reduce energy density due to added volume, necessitating careful thermal design and material selection. Nonetheless, their simplicity, reliability, and proven efficacy make them a valuable component in next-generation BTMS aimed at enhancing safety and suppressing TRP in high-energy-density LIB packs.