blog about Decadelong Data Reveals Flaws in Lithiumion Battery Aging Models

Modern lithium-ion batteries, the powerhouses behind electric vehicles and grid storage systems, face an invisible adversary during their predominantly idle lifetimes. While these batteries spend approximately 90% of their existence in storage—such as parked electric vehicles—they continue to degrade through a process called calendar aging, where parasitic reactions gradually reduce capacity and increase resistance.

Understanding calendar aging presents a unique temporal challenge—significant degradation data at room temperature requires years to collect. Scientists typically circumvent this by gathering data at extreme temperatures over shorter periods, then extrapolating through accelerated aging models. These models traditionally rely on two fundamental principles: the t ^0.5 time dependence (reflecting diffusion-limited growth of the solid electrolyte interface layer) and Arrhenius-type temperature dependence.

While numerous studies initially supported these traditional models across various battery chemistries—including graphite anodes paired with nickel-manganese-cobalt (NMC) or lithium iron phosphate (LFP) cathodes—emerging research reveals significant deviations. Some batteries demonstrate alternative power-law time dependencies (t ^b ), while others maintain Arrhenius temperature behavior but abandon the t ^0.5 relationship. These discrepancies suggest more complex degradation mechanisms at play, potentially involving cathode electrolyte interface growth, transition metal dissolution, or copper current collector corrosion.

Most calendar aging studies span months to five years, yet real-world batteries require decade-long performance. Recent extended-duration studies reveal critical insights:

• Passive anode overhang effects diminish after one year, giving way to linear aging trends
• Power-law exponents change dramatically over time, similar to "knee points" observed in cyclic aging
• Arrhenius deviations only emerge on longer timescales

These findings suggest that models validated with short-term data may significantly misrepresent long-term degradation.

A groundbreaking study analyzing 232 batteries across eight types, four chemistries, and five manufacturers over 13 years reveals several paradigm-shifting conclusions:

• Temperature dependence varies unexpectedly: Significant Arrhenius deviations occur even among similar batteries from the same manufacturer, potentially causing years of prediction error for room-temperature aging.
• Time dependence evolves: The idealized t ^0.5 relationship gives way to less self-passivating values, with substantial variation across chemistries.
• Divergent degradation pathways: Capacity and power fade follow distinct, uncorrelated trajectories.
• Individuality matters: Cell-to-cell variation accounts for substantial degradation differences, emphasizing the need for single-cell analysis alongside population trends.

These findings necessitate a fundamental reevaluation of battery aging models and management strategies. Future research directions should prioritize:

• Developing next-generation aging models incorporating multiple degradation mechanisms and machine learning
• Optimizing battery management systems based on refined understanding of calendar aging
• Improving manufacturing consistency to reduce cell-to-cell variation
• Exploring novel materials and designs with enhanced longevity

As the world transitions toward electrification and renewable energy storage, accurately predicting and mitigating battery aging becomes increasingly crucial. This research provides the foundation for developing more durable, reliable energy storage solutions to power our sustainable future.