
UPS vs Portable Power Station for Home Office Backup
Nine of the first ten results are written by someone selling one of the answers. The number that settles it is in Intel's ATX design guide: 16 ms of PWR_OK hold-up.
What each class of portable station genuinely runs, for how long, and where the capacity goes.

Nine of the first ten results are written by someone selling one of the answers. The number that settles it is in Intel's ATX design guide: 16 ms of PWR_OK hold-up.

Every page on this answers solar panels and stops. Five manufacturers publish the figures that decide it: a solar input ceiling that varies five and a half times, a refill quoted in hours of sunlight rather than hours, and a wall charger whose speed was decided when you bought the box.

Every page answers this with zero emissions and a six-inch rule. We opened four manuals instead: two say the device is for indoor use only, and the only clearance figure anyone publishes is 30 cm, in a troubleshooting table.

Cycles are the wrong unit for a machine that spends its life switched off in a cupboard. At one cycle a month even the worst rating here is forty-two years, and the only commitment anyone makes is two to five.

Using one below freezing and recharging one below freezing are different questions with different answers. Four manufacturer specification pages, and the charging floor does not move.

Everything written about these chemistries is written about cells and then applied to products. Taken to three manufacturer specification pages, all three of the usual conclusions come apart.

The lists of what modified sine wave breaks are usually unsourced. We read what three inverter manufacturers put in their own manuals, and the interesting thing is how little the lists overlap.

Three defensible answers sit behind this question and most pages pick one without saying which. The one written by the people who built the machine says 720 Wh, which this tier does not reach.

The starting figure every page quotes comes from a rule of thumb. The one manufacturer that publishes locked rotor amps disagrees with it on every pump in its catalogue, and by most on the smallest.

The certified rating every runtime calculator uses is measured with the ice maker switched off. NIST measured one running, and at 1,000 Wh it decides whether a twelve-hour night is covered at all.

Doubling the battery doubles the runtime exactly, which is the least interesting thing about this tier. What changes is the surge rating, and whether the total covers the outage the EIA says Americans actually get.

About 800 Wh of a 1,000 Wh station reaches an AC load: 97 hours of a router, 15 of a refrigerator, 32 minutes of a space heater. What it will run at all is decided by the inverter, and three units sold as 1000Wh ship inverters of 1,000, 1,500 and 2,000 W.

Amp-hours are not a measure of capacity. One manufacturer rates a single kilowatt-hour battery at both 19.5 Ah and 312 Ah, and both are correct.

Watts decide whether an appliance starts at all; watt-hours decide how long it runs. This guide sizes a station from sixteen verified appliance articles rather than from runtime claims, and shows why a rated 1,024 Wh delivers about 580 Wh to a small AC load.