State of health for lithium iron phosphate (LFP) storage is usually expressed as retained capacity against beginning-of-life capacity, falling by roughly 1–3% per year in early life and then tapering. Sources describe both calendar and cycle ageing, note that stationary LFP warranty endpoints of 65–70% are contractual conventions, and show that storage duration and duty cycle determine whether capacity or power fade binds first.
How LFP state of health is defined
The dominant definition is capacity-based: present usable capacity divided by rated nameplate capacity under a stated reference, commonly around 0.25C to 0.5C at 25 °C over a defined voltage window. A system delivering 92% of its original energy is at 92% state of health. State of health is distinct from state of charge: state of charge says how full the battery is now, while state of health says how much capacity survives ageing.
State of health is a family of definitions, not one number. A resistance-based variant tracks the rise in DC internal resistance against its beginning-of-life baseline; a power-based variant tracks how hard the system can still charge or discharge within voltage and temperature limits. Vendor battery management system screens often show a proprietary composite index under the same label, so the definition must be checked before comparing figures.
Capacity fade, calendar ageing and cycle ageing
Two mechanisms drive capacity down. Calendar ageing is passive degradation from time, temperature and resting state of charge, so a battery fades even when idle. Cycle ageing is loss of cyclable lithium and active material from throughput, scaling with depth of discharge, C-rate and temperature per cycle. One source gives LFP degradation at about 2–3% per year before augmentation; another describes capacity falling about 1–3% per year in early life and tapering.
Cycle-life ratings sit alongside these annual figures. A quality prismatic LFP cell reaches roughly 4,000–6,000 cycles at 80% depth of discharge under datasheet conditions of 25°C and about 0.5C, according to one source. Another notes that 80% of nameplate is the conventional end-of-life point for cycle-life ratings, while a separate glossary says 65–70% is a contractual convention for stationary LFP, not a physical cliff.
Why duration and duty cycle change the degradation picture
Degradation is not only a capacity question. Resistance growth deepens voltage sag at the same current, so the battery management system trims power limits even while stored energy holds up. In short-duration, high-C-rate duty, power fade can bind before capacity fade does; a gently cycled energy-shifting plant usually hits the capacity limit first. Because utility-scale plants contain hundreds of thousands of cells that age unevenly, the weakest strings tend to set the effective state of health of a rack or container.
Duration also affects the commercial frame around degradation. One source notes that capex per kWh falls and capex per kW rises as duration increases, and that LFP degradation of about 2–3% per year sits alongside fixed operating costs before augmentation. Augmentation is carried as a separate reserve line to offset cell degradation, and retained megawatt-hours anchor warranties, sizing margins and augmentation plans.
What to check in specifications and warranties
Capacity-based state of health is only meaningful against a stated reference: capacity measured at a specified C-rate and temperature, over a defined voltage window, compared with the rated beginning-of-life value under the same conditions. The basis also matters, because a cell-level DC number is not the same as deliverable AC energy at the point of interconnection, which additionally reflects power conversion system, transformer and auxiliary losses.
For resistance-based guarantees, the source warns that measured resistance depends on state of charge, temperature and pulse duration, and that there is no industry-standard endpoint, so any guarantee must define its own reference measurement and allowable growth. Cycle-life claims likewise need depth of discharge, C-rate, temperature and end-of-life threshold; request the capacity-retention-versus-cycle curve and the test standard, such as IEC 62660-1, GB/T 31484, UL 1973 or IEC 61427-1.
Key Takeaways
LFP state of health is usually capacity retained as a percentage of beginning-of-life capacity, not state of charge.
Capacity fade is driven by calendar ageing and cycle ageing, with sources describing roughly 1–3% per year in early life and about 2–3% per year before augmentation.
In short-duration, high-C-rate duty, power fade can bind before capacity fade; energy-shifting plants usually reach the capacity limit first.
Cycle-life and state-of-health figures are only meaningful with stated test conditions and a defined end-of-life threshold.
Sources