What Size Power Station for Emergency Backup

Here’s the trap almost everyone falls into: they pick a power station by matching watt-hours to their daily energy use, plug in their fridge when the lights go out, and discover the unit shuts off immediately. Not because the battery is dead — it barely moved — but because the inverter hit its ceiling. A big battery that can’t run your fridge is just an expensive paperweight.

The rated watt-hours on the box are also not what you get. Tested units deliver less than the nameplate figure, and in a real outage you’re planning against a number that was never accurate to begin with. Size wrong and you find out at the worst possible time. Here’s how to size right.

The Inverter Ceiling Comes Before the Battery

Capacity tells you how long you can run something. The continuous-watt rating tells you whether you can run it at all. That distinction is everything in an emergency.

The loads that matter most during an outage — refrigerators on startup, space heaters, well pumps, microwave ovens — are exactly the loads that stress an inverter hardest. A space heater pulling a steady 1,500 watts will trip a smaller unit outright, not deplete it slowly. A fridge compressor kicks on with a surge that can be two to three times its running draw. If the unit’s continuous or surge rating isn’t high enough, the unit shuts off. No error message, no grace period — just off.

The space heater is the clearest illustration. Testers found that a unit in the roughly 860–1,000 Wh class — which sounds substantial — ran a space heater for about 30 minutes before dying, and that’s only if the continuous rating was high enough to sustain the draw in the first place. Many smaller units can’t hold a 1,500-watt resistive load at all. Resistive heat is the fastest way to find out your unit is undersized.

The practical rule: before you look at watt-hours, find your highest-draw device and check two numbers on the unit’s spec sheet — continuous watts and surge (peak) watts. Your biggest simultaneous load has to fit inside the continuous rating. Any motor-start spike has to fit inside the surge rating. If it doesn’t, the battery capacity is irrelevant.

For reference, manufacturer ratings on tested units look like this:

These are manufacturer figures reported by reviewers, not independently load-tested — treat them as the claimed floor, and check the current spec sheet for whatever model you’re considering.

Rated Watt-Hours Are a Ceiling, Not a Promise

Once you’ve confirmed the inverter can handle your loads, capacity tells you runtime. But the number on the box isn’t the number you get.

When testers have actually measured delivered AC output against rated capacity, the results come in below the nameplate. One measurement found a unit delivering 92% of its listed capacity. OutdoorGearLab’s measured figures show a similar pattern — a unit in the roughly 2,000 Wh class measured 1,710 Wh of usable AC, and a 300-class unit came in at 260 Wh measured versus the roughly 300 Wh nameplate. Inverter conversion losses and a small reserve eat the difference.

This research rests on a handful of data points, not a statistically verified constant — so treat the gap as a realistic planning haircut rather than a precise rule. But the direction is consistent: the real number is lower. If you’re sizing tightly to a specific runtime need, plan against roughly 85–92% of whatever the label says, and know you might land at the low end of that band.

Cold temperatures make this worse. Battery chemistry loses available capacity in the cold, so a unit that’s been sitting in an unheated garage during a winter storm will deliver even less than its warm-weather measured output.

How Long Will It Run a Refrigerator?

The fridge is the anchor for most emergency plans, so it’s worth getting specific — and specific about why generalizing is dangerous.

In actual timed tests, a 1,070 Wh unit ran a 25-cubic-foot refrigerator for 18 hours and 22 minutes. A 268 Wh unit ran a refrigerator for 3 hours and 45 minutes. Those numbers are real, but they’re not directly comparable — different fridges, different sizes, different ambient conditions. The variance is the lesson: fridge runtime swings enormously based on the fridge’s age and efficiency, the ambient temperature, how often the door opens, and whether there’s a freezer compartment cycling alongside.

A rough planning figure of around one hour of runtime per 60–80 usable watt-hours is a starting point, but treat it as directional, not a guarantee. In a hot summer outage where the fridge is opened frequently and the compressor runs almost constantly, you’ll land at the low end. In a mild spring outage with a modern, efficient fridge, you’ll do better.

The 18-hour figure that gets quoted and shared is from one specific test on one specific large, presumably modern fridge. If yours is older, smaller, or the ambient is 85°F, planning around that headline will leave you short.

How Weight Scales With Capacity

Once you’ve worked out the minimum watt-hours and continuous-watt rating you need, weight is the reality check. It scales hard — roughly linearly, with some variation by chemistry and form factor:

• ~300 Wh class: around 7 lbs (Jackery Explorer 300: 7.1 lbs / 260 Wh measured)
• ~1,000 Wh class: around 24 lbs (Jackery Explorer 1000 V2: 23.8 lbs)
• ~2,000 Wh class: roughly 39–53 lbs depending on model (Explorer 2000 V2 at 38.9 lbs, Bluetti Elite 200 V2 at 53 lbs)
• 3,000+ Wh whole-home class: 113–114 lbs before you add anything (EcoFlow Delta Pro 3)

That last tier stops being “portable” in any real sense. And expansion batteries — which add capacity to larger units — contribute roughly 100 lbs each, making the whole system effectively stationary. If your outage plan involves carrying the unit up stairs, across a garage, or into a vehicle, the 2,000 Wh class is approximately where one-person carry ends for most people. Factor this in before you chase maximum watt-hours.

Recharging: Fast From the Wall, Slow From the Sun

Modern LiFePO4 units recharge from wall power quickly — measured examples landed between 65 minutes and about 107 minutes for a full charge. Sources genuinely agree here: sub-two-hour AC charging is now typical for this category. That’s useful for short rolling blackouts or if you have a generator to top up from.

What it doesn’t help with is an extended grid-down event, which is the exact scenario most people are buying for. If the grid is out and solar is your only recharge path, plan for a slow trickle, not a fast refill.

The solar input specs on these units look impressive — rated maximums run from the hundreds of watts up to 2,600 watts on bigger models. Those are manufacturer ceilings under ideal conditions, and every one of those numbers comes from a spec sheet, not a field measurement. Real-world harvest depends on your panel wattage, the angle, cloud cover, time of year, and heat derating on the panels themselves. What you actually get on a partly cloudy day with panels lying flat on the ground is a fraction of the rated max — treat solar as something that extends your runtime over multiple days, not something that refills the battery by afternoon.

One more detail worth knowing: fast-charge mode often needs to be manually enabled in the unit’s app. If you plug in and expect the fastest possible charge rate without checking, you may be on standard mode without realizing it.

Battery Longevity: What the Cycle-Life Claims Actually Mean

LiFePO4 units are marketed with cycle-life figures in the 3,000–6,000 range — one manufacturer claims 6,000 cycles to under 80% capacity on their flagship unit. For typical backup use, where you might run through one full cycle every few months, those numbers represent many years of useful life.

The catch is that no reviewer can verify these figures. Running 6,000 charge cycles would take years of continuous testing — so these are chemistry-based engineering estimates, not independently confirmed results. Treat them as a reasonable expectation for the chemistry, not a tested guarantee. And a naked cycle count without context means less than it sounds: operating temperature, depth of discharge, and charge rate all affect how quickly a cell reaches that 80% threshold. The headline number assumes near-ideal conditions. Deep daily cycling in a hot environment will get you there sooner.

The Sizing Framework

Pull all of this together into a decision sequence:

1. List your must-run devices and their wattage. Focus on what you genuinely need during an outage, not everything you own.
2. Find your highest simultaneous load. This is your continuous-watt floor. Add up everything you’d run at the same time — the fridge compressor while the lights are on, for instance. The unit’s continuous rating must exceed this.
3. Check surge for any motor-load devices. Fridges, pumps, and anything with a compressor surge on startup. The unit’s peak/surge rating must cover that spike.
4. Estimate your daily watt-hour need. Running watts × hours per day, summed across devices.
5. Apply the haircut. Plan against 85–92% of nameplate capacity, and lean toward the lower end if the outage might be cold or extended.
6. Reality-check the weight. A unit that technically meets your energy needs but requires two people to move, or won’t fit where you need it during an outage, is the wrong answer.

The continuous-watt ceiling and the weight are the two constraints that most buyers discover after purchase. The watt-hours are what most buyers shop by. Work the list in the order above, and you’ll end up with a unit that actually runs your essentials when you need it — not one that trips on the fridge and sits there full.