Cycle life

One cycle may be a full 100% DOD charge-discharge or a shallow excursion within a narrow SOC window; the charge throughput and electrode stress differ. A protocol should also state charge and discharge rate, constant-voltage termination, voltage range, temperature, rest, end-of-life capacity, periodic reference tests, and whether the count is raw cycles or equivalent full cycles. “Cycle count” alone cannot distinguish low-rate laboratory life, pulse duty, or a realistic storage schedule. When comparing different protocols, cumulative energy or Ah throughput can help, but efficiency and power change still need retention.

Loss of sodium inventory leaves fewer charge carriers for reversible reaction, loss of active material removes reaction sites, and resistance growth reaches voltage limits earlier under load. High-SOC phase transitions in layered oxides, hard-carbon interphase evolution, cold sodium plating, electrolyte decomposition, and gas may all contribute, but the dominant mechanism changes with chemistry and duty. Teardown and degradation-mode work on commercial NFM-hard-carbon cells in 2025–2026 found active-material loss in some samples and pronounced plating or cell divergence in others. Lifetime improvement therefore begins with diagnosis rather than applying one additive or a lower rate as a universal explanation.

Cells from one lot may still show distributions in initial capacity, resistance, and ageing rate. Some degrade slowly for a long period before accelerating at a knee point. Periodic average-capacity checks can miss early self-discharge, gas, resistance, or outlier-cell changes. Lifetime testing should use sufficient replicates and track capacity, energy efficiency, DC resistance, rest-voltage loss, temperature, and suitable diagnostic signals. Series-connected applications care especially about tail cells because the first cell to reach a limit constrains the string. Results should include sample count and spread rather than the best unit alone.

Calendar ageing depends on storage temperature, SOC, time, chemistry, and interphase condition. High-SOC standby can hold some cathodes at high potential, while cold storage and cold charging are separate problems: the former emphasizes long-term interphase and electrolyte change, the latter also introduces charging polarization and plating risk. Storage, backup, and seasonal equipment may spend much of life at rest, so continuous cycling alone can over- or under-predict service years. A plan should include representative storage points, periodic reference tests, and recovery steps and distinguish reversible recovery, irreversible loss, and power change.

Real equipment combines changing power, temperature, SOC, rest, and occasional extremes rather than repeating one standard cycle. Engineering models can segment duty, accumulate energy throughput and time, then update against field data, but extrapolation retains uncertainty. A change in material, supplier, electrode, formation, software, or cooling may require model recalibration. Product commitments should state end-of-life criteria, capacity and power warranty, permitted duty, and maintenance. For a buyer, stable lots and traceable data are often more valuable than the highest laboratory cycle count.