When you look at a cell datasheet and see a single cycle-life number, someone has already made four decisions on your behalf — decisions that determine whether you’ll see that number in the field or half of it. Before you trust a cycle-life claim, learn to ask for the conditions.

The four hidden variables

A cycle-life claim is a function of four test parameters. Change any of them and the number moves, sometimes dramatically.

1. Depth of discharge (DoD)

Cycling a cell from 100% to 0% stresses it more than cycling from 90% to 20%. Lab data typically uses 100% DoD because it’s the worst case — but most wearables and AR devices never see 100% DoD in the wild. A cell spec’d at 500 cycles / 100% DoD often delivers 900–1200 cycles / 80% DoD and 1800+ cycles / 50% DoD. If your real use case is 70% DoD, multiply the datasheet number by ~1.4 and you’ll be closer to reality.

2. Charge and discharge C-rate

“0.5C / 0.5C” is the polite laboratory standard. A device that fast-charges at 1C and discharges in high-power bursts at 2C will age faster — typically 25–40% faster. Ask the vendor for cycle data at your actual charge rate. If they don’t have it, request a custom aging test (cost: USD 2–5k, duration: 8–16 weeks).

3. Temperature

Cycle life halves roughly every 10 °C increase above 25 °C. A cell that does 500 cycles at 25 °C will do 300–350 at 35 °C and 200 or fewer at 45 °C. Thermally-challenging products (outdoor IoT, charging cradles without airflow, AR glasses in the sun) need to derate cycle life as a first-order design input, not a footnote.

4. End-of-life threshold

Is 80% SOH the end? 70%? 60%? The industry mostly reports 80%, but some medical and automotive specs use 70% because the device can still function usefully there. Always check. A cell rated 500 cycles to 80% typically reaches 700–800 cycles to 70%.

Calendar aging: the one nobody tests

Cells don’t only age by cycling. They age sitting on a shelf, sitting in a warehouse, sitting in a returned device. Calendar aging is driven by temperature and state of charge — a cell stored at 100% SoC and 40 °C loses roughly 10–15% of its capacity in a year even without being cycled.

Two consequences:

  • If your product sits in retail inventory for 6–12 months, calendar aging may be a bigger factor than cycle aging for the first year of customer ownership.
  • Shipping and storing at 30–50% SoC (as DGR requires for air shipping, conveniently) extends calendar life substantially.

What a good cycle-life plot looks like

When you get a cycle-life curve from a vendor, look for four attributes:

  1. Stated conditions. DoD, charge rate, discharge rate, temperature, and end threshold, all legibly marked on the plot.
  2. Sample size. A single cell’s curve is a data point, not a statistic. Fleet averages across 5+ cells with min/max bars are far more useful.
  3. A linear or gently-curved decay, not a knee. If the curve drops sharply below 85% SOH, some internal degradation mechanism is accelerating — you’ll see it as a warranty spike.
  4. Resistance growth data on the same cells. Capacity fade and impedance rise tell different stories. A cell that’s at 85% capacity but with 2.5× internal resistance will feel “tired” to the user long before it hits 80% SOH.

Three questions to ask every vendor

  1. “Can you show me cycle data at our actual charge rate, discharge profile, and ambient temperature, on at least 5 cells?”
  2. “What was the internal resistance at 200 and at 500 cycles?”
  3. “What calendar-aging data do you have at 60% SoC and 40 °C?”

If the vendor can answer all three without qualification, the cycle-life number on the datasheet probably survives field use. If not, plan to run the aging test yourself — or expect surprises.