How we rate power stations

Solar Power Auditor does not currently own, operate or bench-test the products it writes about. We want to say that at the top rather than bury it, because a lot of sites in this category imply hands-on testing they have not done.

What we do instead is narrow and, we think, useful: portable power stations are sold on a rated capacity printed on the box, and they deliver noticeably less than that at the AC outlet. Independent laboratories measure the gap. Manufacturers rarely publish it. We collect those published measurements, put them beside the manufacturer's own figures, and show you both — with a link to whoever did the measuring.

Every measured figure on this site is attributed inline and listed again in a Sources block at the foot of the page. The figure carries a superscript number; the number is the source.

We use three kinds of information, and we never blur them together:

The bulk of the measured capacity data here comes from OutdoorGearLab, which publishes both its results and its procedure, and from StorageReview, which publishes instrumented measurements for some units. Where a unit has never been independently measured, we say so and we do not score it on capacity.

Every scored unit gets one weighted score out of ten, built from the same five criteria in the same proportions. The score box on each review prints the full sum, so you can check our arithmetic.

The weights are deliberate. Usable capacity and output carries 30% because it is the thing most often misrepresented and the thing that decides whether the unit does its job. Value is only 15% because the cheapest unit per watt-hour is frequently the wrong unit for the use case.

Runtimes on this site are calculated, not observed. We take an independently measured usable capacity and divide it by a stated load. Those assumptions are fixed across every page so the comparisons stay honest:

So a unit measured at 890 Wh usable runs the camp fridge for 890 ÷ 45 = 19.8 hours, and we print 19 h 47. Solar recharge divides the rated capacity by the effective panel input, capped by the unit's own solar ceiling — a unit that only accepts 220 W cannot use 2 × 100 W panels any faster than 150 W of real input allows.

These are estimates with the working shown, not measurements. Your fridge, your sun and your cable losses will differ.

• Street prices. They move weekly and we do not have a price feed. We publish the manufacturer's list price with the date we checked it, and nothing else.
• Noise figures we cannot source. If no independent dBA measurement exists, we either quote the manufacturer's claim as a claim, or leave the field empty.
• Runtime for unmeasured units. If nobody has measured a unit's usable capacity, we do not calculate what it will run. We mark it as unmeasured.
• Scores for units nobody has measured. They appear in spec tables but not in rankings.

If a figure here is wrong, we want to fix it and say that we did. Every page carries a modified date, and material corrections are noted on the page itself rather than silently edited. Tell us what we got wrong.

We intend to run our own bench. When we do, measured figures will be labelled as ours, the procedure will be published here in enough detail to be repeated, and third-party figures will stay on the page alongside them rather than being replaced. Until that happens, everything here is somebody else's measurement and our arithmetic — which is exactly why we cite it.