Cycle life numbers are quoted constantly and compared incorrectly almost as often. A cell advertised at 6,000 cycles and one advertised at 3,000 cycles may deliver similar total energy over their lives, or the second may deliver more. The missing information is what a cycle was defined as.
Definitions, stated precisely
State of charge is how full the cell is at this moment, as a percentage of its present usable capacity.
Depth of discharge is how much of that capacity a particular cycle removes. A cycle running from 90 down to 20 per cent state of charge has a depth of discharge of 70 per cent.
Equivalent full cycles normalise partial cycles. Total charge throughput divided by rated capacity gives a count that treats four 25 per cent cycles as one full cycle. This is the basis most storage warranties use.
Energy throughput, in kWh or MWh, is the total energy moved through the battery over its life. For an asset whose revenue comes from moving energy, this is the honest lifetime metric.
The relationship is non-linear, and that is the whole point
Cycle life rises faster than proportionally as depth of discharge falls. Halving the depth typically more than doubles the achievable cycle count.
Worked comparison. Two hypothetical duty profiles on the same 100 kWh pack, both run to 80 per cent capacity retention:
Profile A: 100 per cent depth of discharge, 3,000 cycles achieved. Equivalent full cycles: 3,000. Energy delivered: roughly 300 MWh.
Profile B: 50 per cent depth of discharge, 8,000 cycles achieved. Equivalent full cycles: 4,000. Energy delivered: roughly 400 MWh.
Profile B ran nearly three times as many cycles but delivered only a third more energy, because each cycle moved half as much. The cycle count difference of 5,000 exaggerates the real advantage considerably, and the equivalent full cycle figures tell the truth.
Run the comparison the other way and the trap becomes obvious. If a supplier quotes 8,000 cycles at 50 per cent depth of discharge and a competitor quotes 5,000 at 100 per cent, the competitor's product delivers more total energy despite the smaller headline number.
The physical reason for the non-linearity is that different mechanisms scale differently. Mechanical fatigue from particle expansion scales with the size of the excursion, so shallow cycles cause disproportionately less of it. Film growth scales with time and temperature and continues regardless. Plating risk concentrates at the extremes of the window rather than distributing evenly across it.
The strength of the effect also depends on chemistry. Preger and colleagues, cycling commercial 18650 cells at Sandia National Laboratories, found cycle life strongly sensitive to depth of discharge for NCA and NMC cells, while LFP cells were comparatively insensitive to it and delivered the most equivalent full cycles under nearly every condition tested.
Position in the window matters too
Two cycles of identical depth are not equivalent if they sit in different parts of the range. A cycle between 80 and 100 per cent state of charge ages a cell faster than a cycle between 40 and 60 per cent, despite both having a 20 per cent depth of discharge.
The upper region carries a higher cathode potential, which accelerates electrolyte oxidation, and a lower anode potential, which accelerates film growth and narrows the margin before plating. The very bottom of the range brings its own problems, including higher resistance and, at the extreme, over-discharge risk.
The operational consequence is that a battery management strategy has two levers, not one. Narrowing the window helps. Centring it helps too, and the two are independent. This is why guidance to keep a phone between 20 and 80 per cent is better advice than simply "avoid deep discharges", and why storage operators managing idle state of charge should target the middle rather than merely avoiding the top.
Calendar ageing sets the floor
Restricting depth of discharge has diminishing returns, because a cell also ages while doing nothing.
Consider a stationary asset cycling once daily for twenty years. That is roughly 7,300 equivalent full cycles and also twenty years of calendar exposure. If the cells would have lost a substantial fraction of capacity to calendar ageing alone over that period, then narrowing the cycling window to extend cycle life past the calendar limit buys nothing while giving up usable capacity today.
The design question is therefore where the crossover lies for a given duty. In heavily cycled applications, cycling dominates and the depth of discharge lever is worth pulling hard. In lightly cycled applications such as backup power or capacity market reserve, calendar ageing dominates and the useful levers are temperature and resting state of charge instead.
Typical windows and why
| Application | Typical usable window | Reason |
|---|---|---|
| Consumer electronics | Full range exposed to the user, guidance to limit to 80 per cent | User experience wins; buffers reduce apparent capacity, which sells badly |
| Electric vehicle | Manufacturer buffer at both ends, often several per cent each | Protects warranty exposure and gives reserve; the driver sees the net window as 0 to 100 |
| Grid storage, arbitrage | Wide, close to full usable range | Revenue comes from energy moved, so restricting the window directly reduces earnings |
| Grid storage, frequency response | Narrow, centred, held near a target | Shallow bidirectional cycles around a setpoint |
| Backup and standby | Held high, rarely cycled | Availability is the product, so calendar ageing is the cost of doing business |
| Telecom and off-grid | Moderate, sized to avoid deep cycling | Replacement access is expensive, so life is prioritised over capacity utilisation |
The electric vehicle row explains a question that comes up constantly. The buffer is why a pack described as 82 kWh gross may offer 77 kWh usable, and why the displayed 100 per cent sits below the cells' true full charge. Manufacturers choose the buffer size as a trade between advertised range and warranty exposure, and different manufacturers land in different places.
How to read and write a specification
When reading one, look for depth of discharge, rate, temperature, end-of-life criterion and rest periods alongside any cycle life figure. Convert to equivalent full cycles or total energy throughput before comparing anything.
When writing one, specify the duty you actually expect rather than a standard profile, because the difference between a test protocol and a real duty cycle is where warranty disputes originate. Storage warranties handle this by capping annual throughput and specifying a temperature and state of charge envelope, which is a reasonable model for other applications too.
And be careful with the word cycle in commercial documents. If it is not defined in the document, it will be defined later by whoever is arguing.
Informational and educational content only. Not professional, financial, legal, or engineering advice.