Cycle Life at 90% DoD: Why the Depth You Quote Changes the Number

Cycle life is the most quoted number in battery storage and the least comparable. "6,000 cycles" appears on nearly every LFP datasheet in the market, and it means a different thing on almost every one of them, because the conditions under which the cycle was counted are usually somewhere other than next to the figure.

We recently moved our own published basis to ≥6,000 cycles at 90% depth of discharge, 25 °C. This post explains what that change means, why the depth belongs in the same sentence as the count, and what a buyer should ask when two datasheets show the same number.

Depth of discharge is a stated basis, not a setting

The first thing to be clear about: depth of discharge in a cycle-life rating is not a knob the operator turns. It is the condition under which the test was run.

A cycle-life figure is produced by cycling cells between defined state-of-charge bounds at a defined temperature and rate, until capacity falls to an end-of-life threshold — commonly 80% of initial. Change any of those conditions and you get a different number from the same cell. A 90% DoD cycle discharges more energy per cycle and stresses the electrode more than an 80% DoD cycle, so a cell rated for 6,000 cycles at 90% would show a higher count if retested at 80%.

That is the asymmetry a buyer needs to hold: a deeper stated basis is a stronger claim, not a weaker one. When one vendor says 6,000 cycles at 80% DoD and another says 6,000 cycles at 90%, the second is describing more capability, not less. The number is identical and the products are not.

Our own operating window is published as 90–95% DoD, with the cycle life rated at the 90% end of it.

What the basis does to usable throughput

The practical consequence shows up in throughput, which is what a storage asset is actually selling.

Take the container we build. Nominal capacity is 4,915.2 kWh, and at 90% depth of discharge the usable energy per cycle is:

4,915.2 kWh  ×  0.90  =  4,423.7 kWh per cycle

At the rated 6,000 cycles, lifetime DC throughput is:

4,423.7 kWh  ×  6,000  =  26,542,200 kWh  ≈  26.5 GWh

Now run the same count at an 80% basis. The per-cycle energy drops to 3,932.2 kWh, and 6,000 of those is 23.6 GWh — roughly 11% less lifetime throughput for the same headline cycle count. The rating did not change. The basis did.

This matters directly to levelised cost, because throughput is the denominator. A cost per kWh delivered is computed over the energy the asset actually moves in its life, so a difference in the stated basis moves the economics without moving a single price. We set out the method in our post on levelised cost of storage, and the cost bands themselves are on the BESS cost page.

Temperature belongs in the same sentence

Depth is half the condition. The other half is temperature, and it is omitted even more often.

Our basis is stated at 25 °C, which is the standard laboratory reference. Elevated cell temperature accelerates the side reactions that consume lithium inventory and grow the solid-electrolyte interphase, so a cell held warm ages faster per cycle and also ages while sitting idle. A cycle-life figure with no temperature attached is not a specification; it is a number.

For Indian deployments this is not academic. Ambient conditions across much of the country sit well above the reference for large parts of the year, which is why thermal management is a cycle-life question rather than a comfort one. A liquid-cooled system that holds cells near the reference temperature is protecting the rating; a system that lets them run warm is spending it. We covered the engineering side of that in operating BESS at 48 °C.

There is a second-order effect worth stating: operating rate. A plant sized for energy rather than power runs well below its rated C-rate — a four-hour system draws roughly 0.2 C against a 0.5 C rating — which produces less internal heating and less mechanical stress per cycle. Cycle life at that duty sits inside the rated envelope rather than at its edge.

What the rating does and does not promise

A cycle-life rating is a capability statement about the cells under defined conditions. It is not, by itself, a commercial commitment.

The commitment lives in the warranty, and the two are frequently conflated. A datasheet saying 6,000 cycles and a warranty covering a period and a cycle count are different instruments: the first describes expected behaviour, the second defines what happens if the behaviour is not delivered. Our standard cover is 5 years or 6,000 cycles from site acceptance, whichever comes first, and an extended guarantee to 70% state of health over 20 years is available as a priced option rather than as part of that standard cover.

The variable that connects them is cycles per year. Six thousand cycles at one cycle a day is a little over sixteen years; at two cycles a day it is eight. A duty cycle assumption is therefore doing as much work in a warranty as the cycle count itself, and it is worth pinning before signature rather than after.

Chemistry sets the ceiling the basis is measured against

None of this is an argument that the basis is all that matters. The chemistry sets what is achievable before any condition is stated.

Lithium iron phosphate holds its capacity across deep cycling better than nickel-manganese-cobalt formulations, which is the main reason it has taken the stationary storage market despite lower energy density — density is a vehicle constraint, not a container constraint. We set out that comparison in LFP vs NMC for Indian conditions and why LFP is winning utility storage.

What the basis does is make two LFP products comparable. Without it, every datasheet converges on the same round number and the comparison collapses.

The two questions to ask

When you receive a cycle-life figure, two questions resolve almost all of the ambiguity:

At what depth of discharge? If the answer is 80% and a competing quote says 90%, the two 6,000-cycle claims are roughly 11% apart in delivered energy before anything else is considered.

At what temperature, and what is the end-of-life threshold? 25 °C to 80% remaining capacity is the common reference. A figure quoted at a lower ambient, or to a 70% threshold, is a larger number describing the same cell.

A vendor who answers both without hesitation is quoting from a test report. One who cannot is quoting from a brochure.

Why we published the change

We revised our own stated basis rather than leaving the previous figure in place, and publishing the reasoning matters more than the revision. A specification that changes quietly invites the question of what else moved; one that changes with its basis stated is a specification you can audit.

The full product basis, including the operating window and the certification status, is on the product range. For a configuration sized against a specific duty cycle and ambient, talk to our team.