This article is contributed by Dr. Brian May and Dr. Tal Nagourney of Engineering Systems, Inc.
The global push for electrification—from the smartphone in your pocket to grid-scale energy storage—has made battery safety and reliability vitally important. But what happens when something goes wrong with the batteries that increasingly support the foundation of our everyday lives? Severe consequences: property damage, injury, and even loss of life. When a battery fails, the root cause is rarely a single event, but rather, a cascade of phenomena operating on the atomic level all the way up to the system as a whole. Preventing these failures therefore requires a holistic, multiscale analysis.1
Figure 1: An 18650 cell enters thermal runaway due to fire exposure during an ESi experiment.
Multiscale Devices
Batteries are complex energy storage devices that combine materials and polymers science, electrochemistry, and mechanical and electrical engineering. Their functionality relies on processes spanning ten orders of magnitude:2
- Nanoscale — Charge and discharge requires mobile ionic species to move through the cell and intercalate into the active material’s crystalline lattice structure
- Microscale — Long term stability and cyclability require precise particle sizes and morphology and a stable solid electrolyte interphase (SEI) layer
- Mesoscale — Electrode and separator architecture impacts energy storage kinetics and thermodynamics
- Battery cell level — Proper placement and alignment of electrode windings/layers, separators, and tabs promote safe and reliable energy storage and delivery
- Battery pack level — Battery packs containing one or more cells rely on a battery management system (BMS) to detect issues prior to catastrophic failure
The dimensional scale of battery failures may also span many orders of magnitude. For example, poor alignment of electrode windings at the cell level may result in lithium dendrite formation and growth at the nano and microscales, initiating an internal short circuit at the mesoscale and triggering a thermal runaway event that spreads to the cell, device, and beyond. Performing a single point analysis may exclude critical information. Combining multiple techniques is the key to a holistic root cause failure analysis.
Figure 2: Graphical depiction of the internal structure of an 18650 form factor cell.
Battery Failures Come in Many Shapes and Sizes
Battery failures may be categorized as catastrophic or non-catastrophic.3 Catastrophic failures include thermal runaway that rapidly releases the stored energy as heat and destroys the battery, likely also destroying the physical and chemical features that precipitated the failure. Non-catastrophic failures refer to inadequate performance where the battery is unable to meet its energy storage and delivery specifications; the responsible degradation mechanisms likely remain intact and may be monitored in real time with a BMS or identified in the laboratory with visual, chemical, and electrical analyses.
When investigating either type of failure, it is best to analyze intact exemplar batteries from the same vintage so transient manufacturing defects can be differentiated from post-production treatment. Ideally, these exemplars are made from the same batches of raw materials on the same manufacturing equipment on the same date—this provides the greatest likelihood that a similar defect, if one existed in the failed batteries, is present in the exemplars.
Figure 3: Battery cell defects identified using CT imaging.
Outside-In Approach to Failure Analysis
Regardless of the type of failure, a multimodal investigation should begin non-destructively, following the scientific method. The first steps are to review and analyze all available data. This includes the history of the battery, circumstances of the failure, and information stored in the BMS. The investigation may then proceed to visual inspection, photography, and micrography, beginning with device level concerns such as the battery enclosure and wiring before narrowing the focus.
The battery pack and individual cells may be examined for swelling and evidence of mechanical damage, discoloration, moisture intrusion, and electrolyte leakage. Computed tomography (CT) imaging may be utilized to visualize and memorialize the external and internal 3D structure; features that are hidden or appear visually indistinguishable are readily identified by their radiographic density and can be virtually inspected prior to destructive examination. Proper documentation of all non-destructive steps is critical to memorialize the evidence prior to alteration.
Figure 4: CT image of an 18650 cell after thermal runaway initiated by fire exposure.4
The electrochemical performance of intact cells may be characterized with laboratory instruments at this stage. Equipment such as cell cyclers and vector network analyzers (VNA) allow measurement of numerous properties, including the energy storage capacity, DC and AC resistance, and performance variation within a population of cells. When compared to new exemplars, anomalies in these measurements can be correlated with battery history, defects, and damage. For example, long term charge-discharge cycling provides insight into capacity loss as a function of ambient temperature and charge/discharge C-rates. Meanwhile, electrochemical impedance spectroscopy (EIS) provides insight into the cell materials, internal construction, and ionic mobility.
Once all non-destructive steps are complete, the investigation can proceed to the destructive examination. Batteries and cells may be disassembled in a safe environment, such as an argon-filled glovebox, for direct observation of their internal features. Cell disassembly also allows individual components, such as cathodes, anodes, and separators to be extracted for future analysis, including compositional and morphological characterization.
Figure 5: A glovebox with an inert atmosphere provides a safe environment for cell disassembly.
Once these inspections and tests are complete, the collected data should be analyzed and failure mode hypotheses should be produced using inductive reasoning. These hypotheses can then be tested using deductive reasoning, during which they are compared to all known factual data, both from the investigation at hand and the broader body of scientific knowledge. This testing should be designed to disprove the hypotheses to avoid “confirmation bias,” which occurs when the investigator prioritizes data that support their hypotheses. This is contrary to the scientific method, which states that a hypothesis is only valid if it can withstand rigorous challenges. If a single hypothesis remains valid after suitable testing, it can be selected as the final hypothesis. If not, the previous steps should be repeated as necessary until a valid hypothesis is produced. If this is not possible, the failure cause remains undetermined.
It is important to follow this general flow to perform a scientifically valid failure investigation:
Collect data → examine non-destructively → examine destructively → analyze data → develop hypotheses → test hypotheses → select final hypothesis.5
It is important to approach each step diligently so all available information is collected and considered, as observations made during the initial stages may only make sense in the context of later discoveries. Perhaps a pouch cell enclosure exhibits a small electrolyte leak initiating from its interior—bypassing a thorough exterior examination may lead the investigator to miss such a feature and unknowingly destroy it during the destructive examination.
Figure 6: CT imaging of three 18650 cells shows differences in the number of electrode winding layers and their alignment, which impact performance and safety.
The Power of Complementary Analyses
In addition to the need for following a workflow that maximizes evidence preservation, performing a limited subset of investigative activities may only reveal certain pieces of the puzzle. For example, relying solely on a CT image may reveal an anomalous radiographically dense feature, yet its composition and source may be unclear. Physical teardown alone may miss said feature without first observing it in the high contrast CT image. Combining the strengths of both can identify this feature and its location, then physically isolate and inspect it. This provides the opportunity for further microscopy and chemical analysis to help identify its composition and origin.
Due Diligence as a Risk Mitigation Strategy
Following a multimodal investigation is not only valuable when performing a root cause failure analysis. The same principles apply to proactive analyses when a company exercises due diligence with their products. A thorough evaluation during the prototyping phase can identify potential issues before a product hits the market and can help avoid:
- Personal injury to the consumer
- Property damage from a thermal runaway event
- Product recalls due to safety concerns
- Dissatisfied customers and warranty claims due to poor performance
- Harm to brand reputation that may be difficult to repair
While a holistic, multimodal assessment of batteries may seem daunting, it can save time, money, and effort in the future by identifying risks and mitigating hazards in new products, and understanding and addressing the root cause of failures in the field. Rather than merely identifying “symptoms,” a thorough investigation can reveal their underlying causes, painting a detailed picture of the system from the visible to the microscopic.
References
- P. R. Shearing et al., “Multi Length Scale Microstructural Investigations of a Commercially Available Li-Ion Battery Electrode,” J Electrochem Soc, vol. 159, no. 7, p. A1023, Jul. 2012, doi: 10.1149/2.053207JES.
- Kirby W. Beard, Linden’s Handbook of Batteries, 5th ed. New York: McGraw Hill Professional, 2019.
- C. Mikolajczak, M. Kahn, K. White, and R. T. Long, Lithium-Ion Batteries Hazard and Use Assessment, no. July. in SpringerBriefs in Fire. Boston, MA: Springer US, 2011. doi: 10.1007/978-1-4614-3486-3.
- T. Nagourney, J. Jordan, L. Marsh, D. Scardino, and B. M. May, “The Implications of Post-Fire Physical Features of Cylindrical 18650 Lithium-Ion Battery Cells,” Fire Technol, vol. 57, no. 4, pp. 1707–1722, Jul. 2021, doi: 10.1007/s10694-020-01077-8.
- National Fire Protection Association, NFPA 921 — Guide for Fire and Explosion Investigations, 2024, Section 4.3 Relating Fire Investigation to the Scientific Method. [Online]
