Power Station Cycle Life: What 3,000 or 6,000 Cycles Really Means

By Hicham M. · Updated October 17, 2026

A cycle-life number on a spec sheet is the count of full charge-and-discharge cycles before the battery has faded to a set share of its original capacity. The count gets the attention. The share decides what the count means, and makers don’t all use the same one.

1. What a cycle claim actually says

“4,000 cycles” never means the battery dies at cycle 4,001. It means that after 4,000 full cycles under the maker’s test conditions, the battery should still hold at least a stated share of its original capacity, usually 80% or 70%. Jackery, for example, lists the Explorer 1000 v2 at 4,000 cycles “to 70%+ capacity”. EcoFlow lists the DELTA 3 Plus at 4,000 cycles “to 80% capacity”. Same count, different finish line.

The test conditions matter as much. DJI’s spec page for the Power 500 says its 4,000-cycle figure was measured at 77°F (25°C), recharging in the 270 W standard mode and discharging at 500 W. EcoFlow’s DELTA Pro 3 manual says 80% or better after 4,000 cycles “at 0.5C/0.5C at 25°C (77°F)”, which means charging and discharging over about two hours each. Goal Zero’s Yeti PRO 4000 guide says its 4,000 cycles are at a 0.5C discharge and “climate controlled”. A garage that hits 95°F in August is not that lab.

A cycle also means a full one. Researchers count equivalent full cycles: total energy throughput divided by the battery’s capacity. Two half discharges count as one cycle.

2. Why 70% and 80% claims can’t be compared directly

Across our 54 models, 52 have a published cycle-life figure. Of those, our database records the end-of-life basis as 80% for 34, 70% for 11, and doesn’t record one for 7. 2 models have no published cycle figure at all.

Cycle-life basis by brand
Brand	Models	With a cycle claim	Claimed cycles	Basis recorded
EcoFlow	12	10	3,000–4,000	80% (1 not recorded)
Jackery	14	14	500–6,000	80%, 70%
BLUETTI	10	10	2,500–6,000	80% (2 not recorded)
Anker SOLIX	10	10	3,000–4,000	80% (4 not recorded)
DJI	3	3	4,000	80%, 70%
Goal Zero	5	5	500–4,000	80%

From the model pages’ sources, as of September 2026. “Not recorded” means our database has no basis for that model yet, not that the maker never states one.

Going from 100% to 80% is a 20-point loss. Going to 70% is a 30-point loss. A battery rated 4,000 cycles to 70% is allowed to fade half again as much as one rated 4,000 cycles to 80% before it has used up its claim. Anker makes this point in its own marketing: the comparison table on the SOLIX C2000 Gen 2 page sets its “4,000 Cycles to 80% Capacity” against an unnamed rival’s “4,000 Cycles to 70% Capacity”.

Is there a fair conversion?

Not an exact one. The best public evidence we found is a multi-year study at Sandia National Laboratories that cycled commercial LFP, NMC and NCA cells under different temperatures, depths of discharge and discharge rates. The authors report that the cells “exhibited primarily linear degradation behavior”, with slightly faster fade at the beginning and end. If fade really were a straight line, a claim of N cycles to 70% would put the 80% point at about two-thirds of N:

Cycles to 80% ≈ cycles to 70% × 20 ÷ 30 (straight-line assumption)

For the Explorer 1000 v2 that gives roughly 2,700 cycles to 80%. Treat it as a rough floor, not a conversion. EcoFlow publishes two points for the original DELTA Pro: 3,500 cycles to 80% and 6,500 cycles to 50%. A straight line through the first point would reach 50% at 8,750 cycles, so that battery is expected to fade faster later in life. When fade speeds up late, the 80% point of a 70% claim sits somewhat later than two-thirds. And going the other way, stretching an 80% claim out to 70%, overstates it. Sandia’s authors also found that cells of the same chemistry and size from different makers could differ by more than three times in cycles to 80%, so no formula carries over from one brand’s cells to another’s.

Cell makers use both bases too. EVE Energy’s spec for its LF280N cell rates it at 3,500 cycles or more at 1C/1C and 77°F (25°C) to 80% of initial capacity, while EVE’s spec for a 314 Ah cell gives 8,000 cycles to 70% state of health at 25°C.

3. Chemistry matters more than the brand

Claimed cycle life by chemistry
Chemistry	Models with a claim	Lowest	Median	Highest
LFP (LiFePO4)	46	2,500	4,000	6,000
NMC	6	500	800	1,000

Every claim above 1,000 cycles in our data is for an LFP model. Our 6 NMC models claim 500 to 1,000, and 5 of them are no longer sold.

The lab data points the same way. In the Sandia study, LFP cells lasted 2,500 to 9,000 equivalent full cycles to 80%, against 200 to 2,500 for NMC and 250 to 1,500 for NCA (the LFP figures are partly extrapolated, since most LFP cells hadn’t reached 80% when the study was written up). In the study’s statistical comparison at the same cycle count, LFP cells kept on average 9% more capacity than NMC. Depth of discharge mattered for NMC and NCA (5–10% lower capacity when cycled 0–100% rather than over a narrower range) but had little effect on LFP. Temperature worked in opposite directions: between 59°F and 95°F (15–35°C), LFP faded faster as it got warmer, NMC slower.

Practical upshot for an LFP station: running it all the way down now and then isn’t the battery killer it can be for NMC. Heat is the bigger enemy.

4. Cycles in years

Here are the claims in our database turned into calendar time at one full cycle a day and at one a week.

Claimed cycles converted to years
Claimed cycles	Models	Years at 1 cycle a day	Years at 1 cycle a week
500	2	1.4	10
800	3	2.2	15
1,000	1	2.7	19
2,500	2	6.8	48
3,000	14	8.2	58
3,500	4	9.6	67
4,000	22	11	77
6,000	4	16.4	115

The weekly column is the reason cycle count rarely decides an LFP purchase for outage backup. A station that does a full cycle every week of the year needs 77 years to use up a 4,000-cycle claim, and nobody should plan on that. Batteries also age with time, whether they cycle or not. A 2025 modeling study of LFP cells in storage found the surface layer behind that aging grew about eight times thicker over three years at 131°F (55°C) and 90% charge than at 77°F (25°C) and 10%. DJI puts it plainly for the Power 500: an “approximate lifespan of 10 years”, calculated from “one full charge-discharge cycle nearly every day”.

Daily cycling, such as running a fridge off solar every night, is where the claim starts to matter. At that pace an NMC model rated 500 cycles is near its rating in 1.4 years.

5. What the warranty covers

42 of the 52 models with both figures carry a 5-year warranty. At one full cycle a day, 5 years is 1,825 cycles. Across those 52 models, the warranty runs out after a median 46% of the claimed cycles at that pace, so most of a cycle claim sits outside the warranty.

And warranties don’t promise the claimed capacity. EcoFlow’s US policy excludes “normal wear and tear (including, without limitation, wear and tear of batteries, if applicable)”. Goal Zero’s says batteries “are wearable goods”, and that the warranty “does not guarantee that your battery will function at full capacity for the entire Warranty Period” but at “a reasonable capacity” for its chemistry, use, environment and cycle count. Jackery’s policy adds a condition: the battery cell isn’t covered unless you fully charge it within seven days of purchase and at least once every 6 months. None of the three policies we read puts a capacity percentage in the warranty itself.

6. How to read a claim

Find the percentage. If a page gives cycles with no end-of-life share, check the manual’s spec table. Compare 80% claims with 80% claims.
Check the chemistry. LFP claims in our data run 2,500 to 6,000. A high cycle figure on an NMC station would be unusual.
Size for the faded battery. If you plan to cycle daily, size for 80% (or 70%) of today’s capacity. The power station finder adds a margin you can raise for this; how to size for an outage explains the margin.
Keep it cool and don’t store it full. Store it indoors, follow the maker’s storage charge level, and top it up on the schedule the manual and warranty give.

The highest claims in our data are 6,000 cycles, from Jackery Explorer 1500 v2, BLUETTI Elite 200 V2, BLUETTI Apex 300 and BLUETTI Elite 300. Read the basis on each model page before comparing them with the rest.

Questions

Is a 6,000-cycle power station better than a 4,000-cycle one?

Only if both use the same end-of-life basis and similar test conditions. A 4,000-cycle claim to 80% can mean as much as, or more than, a 6,000-cycle claim to 70%. For occasional outage use, either is far more than you will use.

Can I convert a 70% claim to an 80% one?

Only roughly. If capacity faded in a straight line, cycles to 80% would be about two-thirds of cycles to 70%. Real cells don’t fade perfectly linearly, and cells differ between makers, so treat the result as a ballpark.

Does draining an LFP power station to 0% hurt it?

Less than it hurts NMC. In Sandia’s study, cycling LFP cells over their full range had little effect on capacity compared with narrower ranges, while NMC and NCA cells lost 5–10% more. Heat had a clearer effect on LFP.

Power Station Cycle Life: What 3,000 or 6,000 Cycles Really Means

1. What a cycle claim actually says

2. Why 70% and 80% claims can’t be compared directly

Is there a fair conversion?

3. Chemistry matters more than the brand

4. Cycles in years

5. What the warranty covers

6. How to read a claim

Questions

Sources

Keep reading

Calculators

Runtime calculator

What size do I need?

Solar charging calculator

Real-world runtime simulator

Generator vs power station

Compare models