The global energy transition is accelerating at an unprecedented pace, driven by the urgent need to integrate massive amounts of renewable energy and mitigate climate change. Within this paradigm, lithium-ion battery energy storage systems (BESS) have emerged as a cornerstone technology, providing critical services such as grid frequency regulation, peak shaving, and renewable energy firming. However, the safe operation of these large-scale installations faces severe and escalating challenges. A paramount safety concern is the catastrophic failure initiated or exacerbated by thermal runaway (TR) within individual battery cells. This exothermic, uncontrollable self-heating process is often accompanied by the violent ejection of high-temperature gases, combustible electrolyte vapors, and solid particulates. Among the most dangerous cascading effects of TR is the induction of high-energy electric arcs. Characterized by plasma temperatures exceeding 6000 K—far beyond the melting points of common structural materials like copper, aluminum, and steel—these arcs act as potent ignition sources and can cause severe material damage. They are a core factor in transforming a single-cell incident into a system-wide fire or explosion, as evidenced by numerous high-profile accidents worldwide. This article synthesizes recent research progress on the mechanisms of arc induction by lithium-ion battery thermal runaway and the subsequent hazards, aiming to provide a comprehensive foundation for enhancing the electrical safety design of future energy storage systems.
The formation of an electric arc within a lithium-ion battery system is not a singular event but a complex, multi-path coupling process typically involving the interplay of thermal, electrical, mechanical, and chemical stressors. These pathways can be systematically categorized into three primary mechanistic chains, as illustrated conceptually below. First, Insulation Failure-Induced Arcing occurs when mechanical deformation, vibration, or thermal stress causes structural displacement (e.g., cell swelling, busbar warping), compromising insulating materials and reducing the safe clearance between energized components below a critical breakdown distance. Second, Thermal Runaway Product-Mediated Arcing is driven by the transformation of the local dielectric environment by TR ejecta. High-temperature gases, electrolyte vapors, and solid particles can drastically lower insulation strength and create conductive bridges. Third, Structural Degradation-Evolved Arcing results from the cumulative deterioration of electrical connections over the system’s lifetime, such as contact loosening or corrosion, which can initiate partial discharges that escalate into sustained arcs. Each pathway can trigger a disastrous cascade: “Thermal Runaway → Insulation Failure / Dielectric Alteration / Conductive Channel Formation → Electric Arc → Intensified Fire/Explosion.”
To effectively mitigate these risks, a deep understanding of each mechanism is essential. The following sections detail the current state of knowledge regarding these arc induction pathways, the characteristics and severity of the resulting arc faults, and the computational tools being developed to model these complex phenomena.
1. Mechanisms of Arc Induction in Lithium-Ion Battery Systems
1.1. Insulation Failure as a Primary Trigger
Insulation materials are the first line of defense against unintended electrical contact in a lithium-ion battery pack. These materials, which include structural adhesives, mica papers, thermal insulation pads, and cell blue film, must maintain their dielectric integrity under normal operating conditions and during fault scenarios. However, the extreme thermal environment during a neighboring cell’s thermal runaway presents a severe challenge. Experimental studies involving heating common insulators to temperatures up to 400°C reveal catastrophic degradation. For instance, structural adhesives begin to pyrolyze and soften around 200°C, completely losing their shape and function by 400°C. Other materials like mica paper and insulating foam undergo significant shrinkage and deformation.
The electrical consequence of this thermal assault is a precipitous drop in insulation resistance. Measurements show that the resistivity of these materials can increase by several orders of magnitude upon heating, as summarized in Table 1. This degradation implies two key failure modes: 1) The physical deformation or loss of the insulator creates a reduced gap or direct contact path between live conductors. 2) The charred or decomposed material may itself become semiconductive. When combined with conductive particles from TR (discussed later) or direct contact due to mechanical displacement, the stage is set for a breakdown. Research has specifically highlighted the vulnerability of aluminum laminate film in pouch cells, which can suffer dielectric breakdown at voltages as low as 400V under thermal stress, directly leading to internal short circuits and thermal runaway. Furthermore, simulations of high-voltage battery clusters (e.g., 1500 V systems) indicate that defects in structural adhesive application can create localized high electric field intensities, significantly increasing the risk of insulation failure and subsequent arcing.
| Material | Resistivity at 200°C (Ω) | Resistivity at 300°C (Ω) | Resistivity at 400°C (Ω) | Observation |
|---|---|---|---|---|
| Mica Paper | 3.04 × 109 | 2.91 × 1011 | 5.20 × 1011 | Minor shape change |
| Insulating Foam | 2.94 × 109 | 1.21 × 1011 | 1.59 × 1011 | Shrinkage at 400°C |
| Structural Adhesive | 1.40 × 1012 | Decomposed | Decomposed | Softens at 200°C, melts by 400°C |
| Cell Blue Film | 3.17 × 109 | 1.73 × 1010 | 1.40 × 1011 | Significant shrinkage at 400°C |
1.2. Arc Induction by Ejected Particles and Electrolyte
Thermal runaway of a lithium-ion battery is a violent process that projects a complex mixture of gases, aerosols, and solid particles into the surrounding module space. The solid ejecta, comprising metallic fragments (from current collectors), active material particles, and carbonaceous matter (from the anode and electrolyte decomposition), have been extensively characterized. Studies show the particle size distribution is wide, ranging from sub-micron soot to millimeter-sized metal chunks. The elemental composition is dominated by carbon, oxygen, transition metals (Ni, Co, Mn for NCM cells), and aluminum or copper.
This particulate cloud fundamentally alters the dielectric properties between electrical connections. Recent experimental setups have quantitatively investigated this phenomenon. By collecting TR ejecta and introducing it into a controlled gap between electrodes connected to a variable DC power supply, researchers have mapped the conditions for particle-induced arcing. The critical finding is that the presence of particles drastically reduces the breakdown voltage. For a typical NCM lithium-ion battery’s ejecta, the arc ignition voltage can be as low as 99 ± 5 V for a 1 mm gap, which is merely 1-2% of the pure air breakdown voltage at the same distance. The relationship between critical breakdown voltage (\(U_c\)), electrode gap (\(L\)), and particle size (\(d\)) has been empirically established, showing that \(U_c\) increases with \(L\) and decreases with larger \(d\):
$$U_c \propto L^2 \quad \text{and} \quad U_c \propto \frac{1}{d^n}$$
where \(n\) is a positive exponent. A critical voltage map has been proposed, defining safe operating boundaries. Notably, large metallic fragments (>2 mm), though constituting a small mass fraction (~5-6%) of total ejecta, are particularly hazardous due to their very low electrical resistivity (on the order of milliohms). They can physically bridge gaps or create stable, low-resistance paths that facilitate sustained arcing.
Beyond solid particles, the liquid electrolyte vapor and mist pose a separate but equally severe threat. Electrolyte components like ethylene carbonate (EC) and diethyl carbonate (DEC) have low flash points and, when dispersed, form a combustible fog. Experiments simulating electrolyte leakage across energized terminals reveal a dangerous sequence: initially, the liquid does not conduct but rapidly evaporates and carbonizes under the high local heat from the applied voltage and subsequent micro-discharges. This carbonized residue then forms a conductive bridge, leading to a full, high-energy arc accompanied by intense ignition and temperatures soaring to several thousand degrees Celsius. The electrical conductivity (\(\sigma\)) of fresh electrolyte is relatively low (e.g., ~13 mS/cm), but its thermal decomposition products are highly conductive, completing a dangerous failure chain.
1.3. Arcing at Battery Connections Due to Structural Degradation
A lithium-ion battery system comprises hundreds to thousands of cells interconnected by busbars, cables, and welds. Over the system’s lifecycle, these connections are subjected to thermal cycling, mechanical vibration, and potential corrosion. This can lead to contact loosening, fretting, or breakage of welds. A poor or intermittent connection creates a point of high electrical resistance. When current flows, this point experiences localized Joule heating (\(P = I^2R\)). If the contact separates slightly while current is flowing, an arc can be struck across the resulting micro-gap. This series arc fault is particularly insidious because it can persist within the circuit, causing continuous overheating and damage without necessarily tripping overcurrent protection devices.
Experimental platforms have been built to study this phenomenon, using a movable electrode to simulate a failing connection with a battery terminal. The studies investigate how factors like battery state of charge (SOC), system voltage, circuit current, and gap distance influence the arc’s initiation and severity.
2. Hazard Characteristics and Severity of Induced Arcs
The hazard posed by an arc in a lithium-ion battery system is multi-faceted, involving extreme thermal loading, potential for direct cell ignition, and the propagation of failure. Key experimental findings are summarized below:
2.1. Influence of Battery State of Charge (SOC)
The SOC of a lithium-ion battery at the time of arcing is a critical determinant of the outcome. Higher SOC correlates with greater stored electrochemical energy, which can be released catastrophically if the arc induces an internal short circuit. Experiments on large-format LiFePO4 cells show a distinct progression:
- Low SOC (0-30%): An arc may cause severe local damage, melting terminals and burning insulation, but may not directly trigger full thermal runaway of the cell. However, it can produce enough heat to degrade safety vents and cause electrolyte leakage.
- High SOC (60-100%): Arcing reliably induces catastrophic thermal runaway. The sequence often involves: 1) Arc ignition and stable burning, 2) Ignition of vented flammable gases by the arc, leading to jet fires, 3) Intense heating of the cell casing, 4) Triggering of internal cell failure and violent TR. The time from arc initiation to cell TR decreases significantly with increasing SOC.
The evolution can be broken into stages: I. Pre-arc, II. Arc evolution, III. Arc extinction, IV. Cell thermal failure, and V. Cooling. High-SOC scenarios often omit Stage III, as the arc transitions directly into cell failure.
| SOC Level | Time to Cell Failure | Observed Phenomena | Hazard Severity |
|---|---|---|---|
| 0% | No TR induced | Local melting, insulation fire | Moderate (Fire risk) |
| 30% | Longest (~tens of seconds) | Gas venting, possible late TR | High |
| 60-100% | Short (a few seconds) | Jet fire, violent TR, explosion risk | Severe |
2.2. Influence of Electrical Parameters and Gap Distance
Controlled experiments varying circuit voltage (\(U_{dc}\)), current (\(I_{dc}\)), and electrode separation gap (\(L\)) reveal fundamental relationships:
- Circuit Voltage (\(U_{dc}\)): For a fixed current, a higher system voltage generally leads to a longer time to arc ignition for a given starting gap. This is because a higher voltage can sustain a longer arc column; the initial breakdown may require more time/energy to establish a stable plasma channel from the initial contact point separation.
- Circuit Current (\(I_{dc}\)): A higher current leads to a shorter time to arc ignition and a more powerful arc. The arc voltage (\(U_{arc}\)) during stable burning is relatively constant for a given gap. Therefore, the arc power (\(P_{arc} = U_{arc} \times I_{dc}\)) is directly proportional to the current. A higher-power arc erodes electrodes faster, produces more radiant heat, and is more likely to ignite surrounding materials.
$$P_{arc} \approx U_{arc} \cdot I_{dc}$$
- Gap Distance (\(L\)): Perhaps counter-intuitively, for a fixed voltage and current, a larger initial gap can sometimes lead to a shorter time to stable arcing in a simulated failing connection. The reason is that a larger final gap requires a longer, higher-voltage arc. The system may spend less time in an unstable, high-resistance “glowing contact” phase and transition more quickly to a full breakdown to support the required gap.
These parameters are interconnected. The minimum voltage required to initiate an arc across a gap filled with TR ejecta can be described by an adapted empirical boundary equation, crucial for system design:
$$U_c \ge k \cdot L^2$$
where \(k\) is a constant dependent on the contaminant type (e.g., ~4.08 for specific particle mixtures). For a 400V system, this defines a minimum safe clearance \(L_{min}\) to prevent particle-induced breakdown.
3. Modeling and Simulation of Arcs in Lithium-Ion Battery Systems
Predictive modeling of arc faults is vital for designing safer lithium-ion battery systems, but it remains a formidable challenge due to the multi-physics, multi-scale nature of the problem. Existing modeling approaches can be categorized as follows:
| Model Type | Principle | Advantages | Limitations for BESS Application |
|---|---|---|---|
| Black-box / Empirical (e.g., Cassie, Mayr) | Represents the arc as a time-varying resistance based on energy balance. $$ \frac{1}{g} \frac{dg}{dt} = \frac{1}{\tau} \left( \frac{ui}{P} – 1 \right) $$ where \(g\) is conductance, \(\tau\) time constant, \(u\) voltage, \(i\) current, \(P\) power loss. | Simple, computationally cheap, good for circuit-level studies. | Does not model internal arc physics. Cannot predict ignition from TR conditions. |
| Magnetohydrodynamic (MHD) Models | Solves coupled Navier-Stokes, energy, and Maxwell’s equations for plasma. $$ \nabla \cdot (\rho \mathbf{v}) = 0 $$ $$ \rho (\mathbf{v} \cdot \nabla) \mathbf{v} = -\nabla p + \mathbf{J} \times \mathbf{B} + \nabla \cdot \boldsymbol{\tau} $$ $$ \mathbf{J} = \sigma (\mathbf{E} + \mathbf{v} \times \mathbf{B}) $$ | Captures detailed plasma physics: temperature, flow, magnetic field coupling. Useful for studying established arc behavior. | Extremely computationally expensive. Typically assumes Local Thermal Equilibrium (LTE). Hard to integrate with TR cell models. |
| Microscopic / Particle-in-Cell (PIC) | Tracks individual particle motions in electromagnetic fields. | Can model non-LTE effects, breakdown phases. | Prohibitively complex for system-scale simulation. |
Current research specific to lithium-ion battery systems is in its infancy. Pioneering work has applied MHD-based finite element models to simulate a stable arc between battery terminals, solving for the coupled temperature, magnetic flux density, and fluid flow fields. These simulations provide valuable insights into the extreme conditions during an arc, such as the localization of peak temperature (>6000 K) in the arc core and the formation of high-velocity plasma jets. However, a significant gap exists: there is no coupled model that dynamically simulates the initiation of an arc triggered by the evolving conditions during a cell’s thermal runaway. Such a model would need to integrate the “thermal-electrical-mechanical-chemical” processes: the heating and deformation of components, the release and transport of conductive/combustible ejecta, the degradation of insulation, and the final dielectric breakdown in a changing atmosphere.
4. Conclusion and Future Perspectives
The safety of grid-scale lithium-ion battery energy storage systems is paramount, and the threat posed by thermal runaway-induced electric arcs is severe and multifaceted. This review has consolidated the current understanding of the primary mechanisms:
- Insulation Failure: High temperatures from adjacent cell TR can degrade or destroy insulating materials, reducing clearances and creating direct paths for breakdown.
- Ejecta-Mediated Breakdown: Solid particles and electrolyte vapors ejected during TR dramatically lower the dielectric strength of air, with critical breakdown voltages potentially falling to 1-2% of normal values. Metallic fragments are especially hazardous.
- Connection Faults: Pre-existing or vibration-induced poor connections can evolve into series arc faults, directly heating and potentially igniting the cell.
The hazard severity is strongly influenced by the battery’s SOC and the system’s electrical parameters, with high-SOC, high-current scenarios leading to the most rapid and violent failures. While computational models, particularly MHD-based approaches, can describe the stable arc phase, a critical modeling gap remains in simulating the dynamic, coupled process of arc triggering during a thermal runaway event.
Future research and development must focus on the following frontiers to enable next-generation safe lithium-ion battery system design:
- Dynamic Coupled Experiments: More experiments are needed that replicate real failure sequences, such as a cell undergoing TR while adjacent live terminals are exposed to its ejecta, to capture the transient conditions of arc ignition.
- Advanced Multi-Physics Modeling: The development of an integrated simulation framework is urgently needed. This “digital twin” for arc faults would couple a detailed lithium-ion battery TR model (including ejecta generation) with component thermal/mechanical deformation, contaminant gas/particle dynamics, and a plasma breakdown model. This would allow for virtual testing of pack designs and safety strategies.
- Arc-Resistant Materials and Design: Based on the understanding of mechanisms, new materials with higher thermal stability for insulation and arc-quenching properties should be developed. Module and pack designs need to incorporate guaranteed creepage/clearance distances even under deformation, effective isolation of potential arc sites, and perhaps active arc detection and interruption systems tailored for DC battery systems.
- Standardization and Testing: Safety standards for lithium-ion battery energy storage systems need to evolve to include test protocols that evaluate the system’s resilience to internal arc faults induced by thermal runaway, moving beyond single-cell abuse tests.
By deepening our fundamental understanding of these complex fault cascades and developing predictive tools, we can design lithium-ion battery energy storage systems that are inherently more resilient, helping to ensure the safe and reliable integration of renewable energy into the global power grid.