How we calculate

One calculation drives everything on this site: the battery size we suggest, which stations pass, and how long each would last. Here is all of it, so you can check our work.

Maintained by WhatStaysOn · Last reviewed October 12, 2026

The calculation

For each appliance: average watts = running watts × duty cycle (the share of time it’s actually running). Add them up for the average load. Then for an outage of H hours:

• Energy needed = average load × H
• Usable energy of a station = rated capacity × efficiency × (1 − reserve)
• Runtime = usable energy ÷ (average load × (1 + margin))
• Battery size to look for = energy needed × (1 + margin) ÷ (efficiency × (1 − reserve))

The battery size we show is rounded up (to the next 10Wh, or 50Wh above 1,000Wh). A station passes the energy check only if its runtime is at least H. That’s mathematically the same statement as “its capacity is at least the size to look for,” so the battery size, the verdict and the runtime can never disagree. An earlier version of this planner applied efficiency twice in the filter; we fixed it, and an automated test now sweeps 540 combinations of load, capacity, efficiency, reserve and hours to confirm the three always agree.

Defaults: reserve 10%, extra margin 0%. You can change both under “Fine-tune the numbers.”

Efficiency depends on the load

A station’s AC inverter loses energy in two ways: a proportional conversion loss, and a roughly constant draw just to stay switched on. So small loads get proportionally less of the battery than large ones. We model:

efficiency = average load ÷ (average load ÷ 0.90 + 12W)

That gives about 43% at 10W, 71% at 40W, 76% at 60W, 82% at 110W and 88% at 380W. We chose it to sit within published independent tests: Outdoor Life measured stations delivering 76–92% of rated capacity at about 380W, and 49–77% at about 34W (where our model gives 68%). Some stations did worse at small loads — Outdoor Life’s figures for the BLUETTI AC180 at about 34W were 40–54% — which our cautious case (53% at 34W) reflects. OutdoorGearLab measured a Jackery Explorer 1000 v2 delivering 910Wh of its 1,070Wh. Real stations differ in their idle draw, which manufacturers rarely publish, so this is an estimate.

• Runtime ranges: the range under each result uses a cautious case (88% conversion, 25W overhead) and an optimistic case (92%, 6W). Verdicts use the middle estimate; “Covers it” means the cautious case also lasts, and “Only just” means only the middle one does.
• DC output: if you tick “Run small devices from DC/USB ports,” only the router, phones, laptop and CPAP move to DC (92% conversion, 2W overhead); everything else stays on the AC inverter, and the two are combined. We warn if the DC devices add up to more than 100W, since many DC and USB-C ports are limited to roughly 100–140W.
• Fixed efficiency: you can switch to a single fixed percentage under “Fine-tune.”

The three checks

1. Battery energy — runtime at least as long as the outage.
2. Running power — the station’s continuous rating is at least the total of everything running at once (all running watts added, a cautious assumption).
3. Start-up surge — the station’s surge rating is at least everything running at once plus the largest extra start-up spike of any one appliance (its surge watts minus its running watts). If either side is unknown, the station can’t be marked as passing; it shows as “Check the surge.”

Boost modes (EcoFlow X-Boost, BLUETTI Power Lifting) run some heating appliances at reduced power. They are never used as a surge rating. Anker’s SurgePad is its name for the C1000’s actual surge capability.

Verdicts

• Covers it — passes all three checks, and lasts the outage even on the cautious efficiency case.
• Only just — passes on the expected case but not the cautious one.
• Check the surge — passes energy and running power, but a surge figure is unknown (usually an appliance you added without start-up watts).
• Short alone — needs its add-on battery — fails only on energy, and would pass at the maker’s maximum documented expansion, with the same output and surge. We only count expansion batteries the maker lists as available.
• Runs out early — fails on energy, with no expansion that fixes it.
• Can’t run this load — fails on running power or surge.

For medical equipment, passing verdicts are relabeled “Estimate covers it — test first.” Discontinued stations are left out of planner results. Results are ordered by verdict, then smallest-that-works first. Whether a brand has an affiliate program plays no part, and we have no affiliate links.

Appliance figures

Product evidence rules

• Specs come from the manufacturer’s US product page or user manual. Where those aren’t available, we use the manufacturer’s other regional sites or a manual hosted by a third party, and say so on the station’s page.
• Where official sources disagree on a figure that affects a verdict, we use the more cautious one and explain the conflict (for example, the BLUETTI AC200L’s surge).
• Unknown is shown as unknown. We never fill a gap with a guess or with a boost-mode number.
• We record availability and warranty as stated on the manufacturer’s US site on the check date. We don’t show prices.
• We have not hands-on tested any station. Where independent tests exist, we cite them as independent tests, not as our own.

Solar and expansion views

These are shown in a station’s details but never change its verdict. Solar: daily energy ≈ the smaller of your panel watts and the station’s solar input limit × good sun hours × 0.75 (covering heat, angle, wiring and controller losses — an estimate; NREL’s PVWatts tool uses a 14% default system loss before other factors). Expansion: we re-run the energy check at the maximum capacity the maker documents.

Corrections log

• Oct 12, 2026 — BLUETTI AC180 surge: changed from unknown to 2,700W after confirming “Surge Power: 2700W” in BLUETTI’s US spec table and “can handle a surge of 2700W” in the user manual.
• Oct 12, 2026 — BLUETTI AC200L surge: BLUETTI’s US page says 4,800W, but the manual’s overload protection trips above 3,600W. We use 3,600W.
• Oct 12, 2026 — BLUETTI EB3A: marked discontinued (per BLUETTI US); removed from planner results.
• Oct 12, 2026 — CPAP default: lowered from 60W to 40W, based on ResMed’s published average of about 34W with heated humidifier and tube.
• Oct 12, 2026 — efficiency: replaced a fixed 85% with the load-aware model above, after independent tests showed much lower delivered energy at small loads.
• Oct 12, 2026 — surge check: fixed how start-up spikes combine. The old rule took the appliance with the highest surge watts plus everyone else’s running watts, which could understate the peak when a high-draw appliance like a heater was on alongside a fridge. It now adds the largest extra start-up spike to everything running.

Found an error? Email [email protected] with the page and, ideally, the manufacturer source. See how corrections work.