What a CPAP actually draws
The figure that matters is not the machine's peak rating but its average overnight draw, and it changes enormously with one setting.
My own machine is at the light end. Two nights on the Anker SOLIX C1000 Gen 2 took its display from full to 58% — roughly 370 watt-hours of its 884 Wh measured usable, or about 185 Wh a night. That's an average of around 23 W over eight hours. Yours may well draw more, which is why the table below gives calculated nights at the typical 35 W and 70 W figures alongside what I actually saw.
The humidifier is a heating element. Turning it on roughly doubles consumption and can more than halve your nights per charge.
Nights per charge, and how loud each was
Usable capacity is my own measurement: each unit drained from full to shutdown on a constant 300 W load, metered at the AC outlet.
| Unit | My measured usable | My CPAP rating | Nights I ran | Calc. nights, 35 W | Calc. nights, 70 W | Noise at ~500 W |
|---|---|---|---|---|---|---|
| Bluetti Elite 100 V2 My pick | 892 Wh | 5/5 | 3, barely audible | 3.2 | 1.6 | 34 dBA |
| Jackery Explorer 2000 v2 Runner-up | 1,702 Wh | 5/5 | 4 | 6.1 | 3.0 | 42 dBA |
| EcoFlow Delta Pro 3 | 3,772 Wh | 5/5 | — | 13.5 | 6.7 | 41 dBA |
| Bluetti Elite 200 V2 | 1,902 Wh | 4/5 | — | 6.8 | 3.4 | 40 dBA |
| Anker SOLIX F3800 Plus | 3,448 Wh | 4/5 | — | 12.3 | 6.2 | 42 dBA |
| Anker SOLIX C1000 Gen 2 | 884 Wh | 4/5 | 2, 58% left | 3.2 | 1.6 | 41 dBA |
| Jackery Explorer 1000 v2 | 905 Wh | 4/5 | 2 | 3.2 | 1.6 | 44 dBA |
| EcoFlow Delta 3 | 862 Wh | 4/5 | — | 3.1 | 1.5 | 43 dBA |
| EcoFlow River 2 Pro | 615 Wh | 4/5 | 1, humidifier off | 2.2 | 1.1 | 39 dBA |
| Bluetti AC70 | 684 Wh | 3/5 | — | 2.4 | 1.2 | 45 dBA |
| Anker SOLIX C300 | 232 Wh | 3/5 | 1, humidifier off | 0.8 | 0.4 | Not tested at 500 W |
Swipe for 3 more columns →
- My measured usable
- 892 Wh
- My CPAP rating
- 5/5
- Nights I ran
- 3, barely audible
- Calc. nights, 35 W
- 3.2
- Calc. nights, 70 W
- 1.6
- Noise at ~500 W
- 34 dBA
- My measured usable
- 1,702 Wh
- My CPAP rating
- 5/5
- Nights I ran
- 4
- Calc. nights, 35 W
- 6.1
- Calc. nights, 70 W
- 3.0
- Noise at ~500 W
- 42 dBA
- My measured usable
- 3,772 Wh
- My CPAP rating
- 5/5
- Nights I ran
- —
- Calc. nights, 35 W
- 13.5
- Calc. nights, 70 W
- 6.7
- Noise at ~500 W
- 41 dBA
- My measured usable
- 1,902 Wh
- My CPAP rating
- 4/5
- Nights I ran
- —
- Calc. nights, 35 W
- 6.8
- Calc. nights, 70 W
- 3.4
- Noise at ~500 W
- 40 dBA
- My measured usable
- 3,448 Wh
- My CPAP rating
- 4/5
- Nights I ran
- —
- Calc. nights, 35 W
- 12.3
- Calc. nights, 70 W
- 6.2
- Noise at ~500 W
- 42 dBA
- My measured usable
- 884 Wh
- My CPAP rating
- 4/5
- Nights I ran
- 2, 58% left
- Calc. nights, 35 W
- 3.2
- Calc. nights, 70 W
- 1.6
- Noise at ~500 W
- 41 dBA
- My measured usable
- 905 Wh
- My CPAP rating
- 4/5
- Nights I ran
- 2
- Calc. nights, 35 W
- 3.2
- Calc. nights, 70 W
- 1.6
- Noise at ~500 W
- 44 dBA
- My measured usable
- 862 Wh
- My CPAP rating
- 4/5
- Nights I ran
- —
- Calc. nights, 35 W
- 3.1
- Calc. nights, 70 W
- 1.5
- Noise at ~500 W
- 43 dBA
- My measured usable
- 615 Wh
- My CPAP rating
- 4/5
- Nights I ran
- 1, humidifier off
- Calc. nights, 35 W
- 2.2
- Calc. nights, 70 W
- 1.1
- Noise at ~500 W
- 39 dBA
- My measured usable
- 684 Wh
- My CPAP rating
- 3/5
- Nights I ran
- —
- Calc. nights, 35 W
- 2.4
- Calc. nights, 70 W
- 1.2
- Noise at ~500 W
- 45 dBA
- My measured usable
- 232 Wh
- My CPAP rating
- 3/5
- Nights I ran
- 1, humidifier off
- Calc. nights, 35 W
- 0.8
- Calc. nights, 70 W
- 0.4
- Noise at ~500 W
- Not tested at 500 W
Calculated nights assume eight hours at 35 W or 70 W — assumptions here. A dash means I rated the unit for CPAP use but didn't log a night count. Noise at about 500 W is far above a CPAP's draw, so treat it as a ceiling on how loud each fan gets, not what you'll hear at 23 W.
1 · Bluetti Elite 100 V2 — three nights, and silent
1:1
Bluetti Elite 100 V2
Calculated, not observed: 892 Wh measured on my bench ÷ 45 W average draw; solar from 150 W effective input (2×100 W derated to 75%). Assumptions.
usable Wh
- Quietest of the eleven: 34 dBA at ~500 W, 47 dBA at full output
- 892 Wh measured — the most of any 1,024 Wh unit on my bench
- Accepted 940 W of solar, far beyond its class
- Beat its own recharge claim: 1 h 12 against a stated 1.8 h
I used the Elite 100 V2 for 37 days, including three nights running my CPAP in the bedroom, and I never noticed it. That is the single most important thing about a CPAP battery after capacity, and it's why I rated noise 5 out of 5 — the only unit of the eleven to get full marks.
On my meter it delivered 892 Wh — 87.1% of its rating. It switched to battery in 10 ms, so with the CPAP plugged into it and it plugged into the wall, an outage passes without interruption. The screen is big and bright enough to read at a glance in the dark.
The catches: a 72-minute wall recharge, slower than the Anker and EcoFlow, and at $899 list it is expensive per watt-hour. If capacity is the only thing that matters to you, the Anker below delivers the same nights for much less.
2 · Jackery Explorer 2000 v2 — if you need the humidifier
1:1
Jackery Explorer 2000 v2
Calculated, not observed: 1702 Wh measured on my bench ÷ 45 W average draw; solar from 150 W effective input (2×100 W derated to 75%). Assumptions.
usable Wh
- 43 usable watt-hours per pound, the best carry ratio of the eleven
- 1,702 Wh measured usable
- Recharge claim held up: 1 h 44 against a stated 1 h 45
- 4,300 W overload trip, the highest in its class
It ran my CPAP for four nights, and 1,702 measured watt-hours works out at about six nights at 35 W or three at 70 W with the humidifier running. If you live somewhere outages run for days, or the therapy isn't optional, this is the capacity that turns a worry into a non-event.
It was quiet in normal use and I could lift it alone at 39.5 lb. Its 400 W solar ceiling and 70% cycle rating are weaknesses in general use, but for a unit that sits charged waiting for an emergency, neither matters much.[2] UPS switchover measured 20 ms.
3 · Anker SOLIX C1000 Gen 2 — the value option
1:1
Anker SOLIX C1000 Gen 2
Calculated, not observed: 884 Wh measured on my bench ÷ 45 W average draw; solar from 150 W effective input (2×100 W derated to 75%). Assumptions.
usable Wh
- Highest measured score of the eleven at 8.54
- 884 Wh delivered on my bench — 86.3% of rated
- Fastest wall recharge I timed at 52 minutes
- 2,000 W continuous, the most in its class
- No expansion battery at all[3]
- Fan and app both rated 3/5 — the weakest parts of living with it
- 588 W of solar accepted against a 600 W ceiling
- 52 dBA at full output, louder than the Bluetti
Two nights left it at 58%, so three nights dry is comfortably within reach — the same capacity as the Bluetti for $400 less at list. It refilled in 52 minutes, so a brief return of mains power tops it up completely, and its 10 ms switchover kept my desktop PC running.
It is quiet below 500 W, which covers a CPAP. Just don't fast charge it beside the bed — the fans roar at around 1,600 W of input.
Practical points
Turn the humidifier off on battery. It is the single biggest lever you have — it roughly doubles your nights.
Use the DC output if your machine supports it. Many CPAPs accept 12 V directly. Skipping the inverter avoids the AC conversion loss and the inverter's idle draw, which I measured at between 7 W and 38 W across these units — at the top end, that's more than the CPAP itself. You'll need the correct DC cable from your machine's manufacturer; don't improvise one.
Keep it charged. A power station sitting at 40% is not a backup. If it's there for medical use, top it up monthly.
Check your own machine's draw. My 23 W average is lighter than typical. Your machine's label or manual gives its rating, and pressure settings move it.
Use UPS mode. I measured switchover times from 10 ms to 30 ms. Leaving the machine permanently plugged into the power station, with the power station plugged into the wall, means an outage causes no interruption at all.
FAQ
How many nights will a power station run a CPAP?
In my testing, two to three nights on a 1 kWh unit with the humidifier off and four on a 2 kWh unit. Calculated at a typical 35 W, a 1 kWh unit gives about three nights dry or a night and a half with the humidifier on.
How much power does a CPAP use?
Typically about 35 W with the humidifier off and 70 W with it on, with heated tubing adding more. My ResMed averaged about 23 W with the humidifier off — around 185 Wh a night. Check your own machine.
Does turning off the humidifier really matter?
It's the largest single factor. The humidifier is a heating element, and switching it off roughly halves consumption.
Will the Anker SOLIX C300 run my CPAP overnight?
It ran mine for one night with the humidifier off, but my machine is light. At a typical 35 W, its 232 measured watt-hours give about six and a half hours — short of a night.
Can I use a power station as a UPS for my CPAP?
Yes. Every unit I tested switched over in 30 ms or less, and the Bluetti Elite 100 V2, Anker C1000 Gen 2, EcoFlow Delta 3 and Delta Pro 3 all switched in 10 ms. Confirm compatibility with your equipment supplier.
Which power station is quietest overnight?
The Bluetti Elite 100 V2 — 34 dBA at about 500 W at one metre, against a 32 dBA room. It was the only unit I never noticed overnight.
Quiet first, then three nights
For a CPAP, my pick is the Bluetti Elite 100 V2: three nights from one charge, a 10 ms switchover, and a fan I never heard. That pick is my opinion; the measurements beside it aren't.
If you need the humidifier every night, or outages where you live run for days, step up to the Jackery Explorer 2000 v2 and its four nights.
If capacity matters more than silence, the Anker SOLIX C1000 Gen 2 does the same nights for $400 less — just don't fast charge it beside the bed.
And don't rely on the Anker SOLIX C300 for this unless you know your machine draws very little.
Sources
- Bluetti. Elite 100 V2 — specifications. Manufacturer spec · accessed 5 September 2026
- Jackery. Explorer 2000 v2 — specifications. Manufacturer spec · accessed 5 September 2026
- Anker SOLIX. SOLIX C1000 Gen 2 — specifications. Manufacturer spec · accessed 5 September 2026