What we read on a spec sheet
Brands print the numbers that look best. These are the ones we check, and the fine print we look for.
Rated capacity vs usable capacity
The watt-hours on the box are the battery's stored energy. Less than that reaches an AC outlet, because the inverter loses some converting DC to AC and the electronics draw power of their own. Jackery's own runtime formula takes off 15% for this (Jackery). We show rated Wh for every model and, where an independent tester has measured it, the measured figure in its own row (see Independent measurements below).
Chemistry and cycle rating
Current EcoFlow, Jackery, Anker SOLIX and Bluetti stations use LFP (lithium iron phosphate) cells. Older and some lighter units use NMC, which carries much lower cycle ratings: Jackery rates its original Explorer 1000 at 800 cycles to 80% and the LFP Explorer 1000 v2 at 4,000 cycles to 70% (Jackery).
A cycle rating means nothing without its end point. "4,000 cycles to 70%" is a weaker claim than "4,000 cycles to 80%", so we always print the percentage beside the count. When a brand leaves it out (Bluetti's AC180 page says "3500+ Cycles" with no threshold, Bluetti), we write "threshold not stated". Cycle counts are the brand's own lab figures; EcoFlow's page says its results are "based on our internal testing" (EcoFlow). We call them "rated", never "lasts".
Continuous, surge and boost output
Continuous output is what the inverter holds for as long as the battery lasts. Surge is a short burst for starting motors, such as a fridge compressor or a sump pump. Boost modes (EcoFlow X-Boost, Anker SurgePad, Bluetti Power Lifting) lower the voltage so a heating appliance draws less. They are not a bigger inverter, and they don't help motors or electronics (EcoFlow, Anker). We list the three figures in separate columns, never add them together, and only quote a boost figure as "resistive loads only".
UPS / EPS switchover
Most power stations are not true online UPS units. When the grid drops, they switch from pass-through to their own inverter, and the gap is the switchover time. Brands claim figures such as "<10 ms" (EcoFlow DELTA 3 Plus) and "≤20 ms" (Jackery Explorer 1000 v2) (EcoFlow, Jackery). For a desktop PC, the power supply has to ride through that gap. The ATX 3.1 standard requires 12 ms of hold-up at full load (Seasonic), so a 20 ms switchover can reboot a heavily loaded PC. We print the claimed figure, or "not published". For a CPAP, check with the device maker or your clinician, as Anker's own guide advises (Anker).
Solar input
The spec that decides which panels work is the input voltage window, not just the watts. Panel voltage rises in the cold, so two panels in series that fit on a mild day can exceed the limit on a freezing morning (EC&M on NEC 690.7). Panels rarely deliver their label rating; NREL's PVWatts model starts from a 14% system loss (NREL). We show the brand's max solar watts and voltage range, and we treat "solar charge in X hours" claims as best case.
Expandability and 240 V
Extra batteries add watt-hours, not inverter watts. Every brand's expansion batteries are sold for its own series, so expanding later means staying with that brand. We list the maximum capacity the brand publishes and which batteries fit.
US homes are wired for 120/240 V. Most portable stations put out 120 V only, and a 240 V load such as many well pumps can't run on 120 V at any wattage. We mark 240 V output as yes or no.
Certifications, and who listed them
A safety listing means an accredited lab has certified the product to a named standard. For portable power stations the product standard is UL 2743. A stationary home battery system is covered by UL 9540; UL 9540A is a fire test method, not a certification (UL Code Authority, IAEI).
UL is not the only lab that can list to a UL standard. Any lab on OSHA's list of Nationally Recognized Testing Laboratories can, including Intertek, CSA Group, SGS North America and TÜV (OSHA). So we look for three things: the standard number, the lab, and a listing we can find in that lab's directory.
What we see on brand pages often falls short of that. A page might say "UL" with no standard number, list component standards instead of UL 2743, or describe a performance rating as a certification. We report it the way we found it:
- Listed: "Listed to UL 2743 by [lab], checked [date]", when we found the listing.
- Brand claim, not verified: when the brand states a standard but we could not confirm a listing.
- Not stated: when the brand names no standard.
One more point from UL itself: several individually certified portable packs wired together into a home system do not make a certified (listed) system (UL Code Authority).
Independent measurements
No station delivers its full rated capacity at the outlet. When a lab has measured a model, we show the measured figure in its own row, beside the rated one. We never mix the two in one comparison.
| Model | Rated | Measured AC | Usable |
|---|---|---|---|
| EcoFlow DELTA Pro 3 | 4,096 Wh | 3,790 Wh | 93% |
| Anker SOLIX C1000 Gen 2 | 1,024 Wh | 890 Wh | 87% |
| Jackery Explorer 2000 v2 | 2,042 Wh | 1,710 Wh | 84% |
Source: measured by OutdoorGearLab with a light-bulb load bank, page updated Jul 23, 2026. Low loads waste more, because the inverter's own draw is a bigger share.
Owner-measured figures. Some of the most useful numbers, such as how much a station draws with its AC outlets on and nothing plugged in, come from owners with a plug-in meter. When we use one, we tag it OWNERS, name where it was posted (for example, "owner-reported on DIY Solar Forum") and link the thread. We don't present owner reports as lab results, and we don't present anyone's measurement as our own.