Evidence summary
Evidence summary
Interpret capacity retention and coulombic efficiency only with the cycling protocol, nominal and measured capacity definitions, temperature, current rate, voltage limits, rest steps and formation history. Compare per-cycle trajectories, early-cycle behavior and cell-to-cell variation rather than one end-of-test number.
Key takeaways
- 01Define the capacity denominator and end-of-life rule before calculating retention.
- 02Plot charge and discharge capacity with coulombic efficiency; do not use CE alone.
- 03Preserve current, voltage, time and temperature traces behind cycle summaries.
- 04Use multiple cells and show cell-level variation before claiming improved lifetime.
Define the cycle and capacity basis
State whether capacity is charge or discharge, gravimetric or areal, and normalized to nominal, formation or maximum measured capacity. Identify partial cycles and diagnostic cycles separately.
- Record voltage limits and C-rate.
- Preserve formation cycles.
- Define end of life before analysis.
Read retention and efficiency together
Coulombic efficiency can reveal small per-cycle imbalance, but capacity trajectory shows the accumulated outcome. Inspect both alongside energy efficiency and voltage profiles when available.
- Do not hide early-cycle transients.
- Flag protocol transitions.
- Compare charge and discharge separately.
Keep cell-to-cell variation visible
A mean line can hide outliers and divergent degradation modes. Show each cell, group summaries and the number of valid cells at each cycle.
- Use consistent censoring rules.
- Report missing or failed channels.
- Separate exploratory and validation cells.
Use prediction only inside its evidence boundary
Early-cycle features can support forecasting, but the training chemistry, protocol and end-of-life definition determine where a model may be trusted.
Methodology and scope
The workflow emphasizes protocol-aware descriptive analysis before predictive modeling. It uses published battery-lifetime studies to show why early-cycle features can matter while avoiding claims that one model generalizes across chemistries and conditions.
Limitations
- Capacity retention is protocol-dependent and cannot be compared across arbitrary rates or voltage windows.
- High coulombic efficiency can coexist with gradual inventory or active-material loss.
- A small cell sample does not establish manufacturing variability or population lifetime.
- Early-cycle prediction models may fail outside their training chemistry and protocol.
References
- [1]
Data-driven prediction of battery cycle life before capacity degradation
Severson et al.. Nature Energy (2019).
doi:10.1038/s41560-019-0356-8 ↗ - [2]
Degradation diagnostics for lithium ion cells
Birkl et al.. Journal of Power Sources (2017).
doi:10.1016/j.jpowsour.2016.12.011 ↗
Suggested citation
Suggested citation
SciPhys Research Team. “Battery Capacity Fade Analysis.” SciPhys, August 5, 2026. https://www.sciphys.com/blog/battery-capacity-fade-coulombic-efficiency
Apply the workflow
Inspect capacity fade with the full cycling context.
Upload battery cycling data with the technique preselected and compare capacity, efficiency and protocol-aware trends.