Figures here come from third-party laboratory tests, each one cited. We earn no commission from any link on this site. How we're funded
Guide · home backup

Running a Fridge Through a Power Outage — What Each Battery Size Actually Buys

A closed refrigerator holds a safe temperature for about four hours on its own. Everything past that is what you are buying a battery for.

Updated 5 September 2026 6 min read Third-party test data

The most common reason people buy a power station is a freezer full of food and a memory of a bad week.

So this guide answers one question precisely: how many hours of refrigeration does each size of battery actually give you, calculated from measured capacity rather than the rating on the box.

First, how long you have without one

Before buying anything, know the baseline. US food-safety guidance is that a refrigerator holds a safe temperature for about 4 hours if you keep the door closed, and a full freezer for about 48 hours (24 hours if half full).

◆ Most
If your outages are typically an hour or two, a power station is not solving a refrigeration problem — the closed door already does. The case for a battery starts at outages that run past about four hours, and becomes compelling past twelve.

The single most effective thing you can do during an outage is not open the door. That is free, and it beats any battery.

What a fridge actually draws

Not what the label says. A refrigerator's compressor cycles on and off, so the figure that matters is the running average.

~150 W Rated draw
~35% Duty cycle
~52 W Running average

A typical full-size household fridge rated at 150 W therefore consumes roughly 52 watt-hours per hour. A 12 V compressor camp fridge averages about 45 W.

Two caveats. Duty cycle rises in a warm house — during a summer outage with no air conditioning, expect closer to 45–50% and size accordingly. And the compressor draws several hundred watts for a fraction of a second when it starts, so the unit's surge rating matters even though the running draw is small.

Hours per battery size

Calculated from independently measured usable capacity at a 52 W average — assumptions here.

Refrigeration runtime by unit
UnitMeasured usableHousehold fridgeCamp fridgeList
EcoFlow Delta Pro 3 Longest3,790 Wh72 h 1184 h 13$3,699
Bluetti Elite 200 V21,910 Wh36 h 2342 h 27$1,199
Jackery Explorer 2000 v21,710 Wh32 h 3438 h 00$999
Jackery Explorer 1000 v2910 Wh17 h 2020 h 13$559
Anker SOLIX C1000 Gen 2 Best value890 Wh16 h 5719 h 47$500
Bluetti Elite 100 V2890 Wh16 h 5719 h 47$899
EcoFlow Delta 3870 Wh16 h 3419 h 20$699
Bluetti AC70690 Wh13 h 0915 h 20$359
EcoFlow River 2 Pro620 Wh11 h 4913 h 47$529
Anker SOLIX C300230 Wh4 h 235 h 07$330

Swipe for 1 more column →

EcoFlow Delta Pro 3Longest
Measured usable
3,790 Wh
Household fridge
72 h 11
Camp fridge
84 h 13
List
$3,699
Bluetti Elite 200 V2
Measured usable
1,910 Wh
Household fridge
36 h 23
Camp fridge
42 h 27
List
$1,199
Jackery Explorer 2000 v2
Measured usable
1,710 Wh
Household fridge
32 h 34
Camp fridge
38 h 00
List
$999
Jackery Explorer 1000 v2
Measured usable
910 Wh
Household fridge
17 h 20
Camp fridge
20 h 13
List
$559
Anker SOLIX C1000 Gen 2Best value
Measured usable
890 Wh
Household fridge
16 h 57
Camp fridge
19 h 47
List
$500
Bluetti Elite 100 V2
Measured usable
890 Wh
Household fridge
16 h 57
Camp fridge
19 h 47
List
$899
EcoFlow Delta 3
Measured usable
870 Wh
Household fridge
16 h 34
Camp fridge
19 h 20
List
$699
Bluetti AC70
Measured usable
690 Wh
Household fridge
13 h 09
Camp fridge
15 h 20
List
$359
EcoFlow River 2 Pro
Measured usable
620 Wh
Household fridge
11 h 49
Camp fridge
13 h 47
List
$529
Anker SOLIX C300
Measured usable
230 Wh
Household fridge
4 h 23
Camp fridge
5 h 07
List
$330
⚠ The
230 usable watt-hours gives about four and a half hours of refrigeration — roughly the same as simply keeping the door shut. For this specific job it is not worth its $330.[1]

Matching a battery to your outages

If your outages lastYou need aboutWhich is
Under 4 hoursNothingA closed door covers it
4–12 hours400–650 Wh usableBluetti AC70 class
Overnight, 12–18 hours650–950 Wh usable1 kWh class
A full day and a half1,700–1,900 Wh usable2 kWh class
Two to three days3,700 Wh+ usableDelta Pro 3 class

Note that these are usable figures, not ratings. To convert to a rating to shop by, divide by about 0.85 — or better, use a unit with a published measurement.

The mistake: sizing for the fridge alone

The fridge is rarely the only thing you want during an outage. Add the realistic rest of it and the numbers change sharply:

Load setAverage drawHours on 890 WhHours on 3,790 Wh
Fridge only52 W16 h 5772 h 11
Fridge + lights82 W10 h 5146 h 13
Fridge + lights + router + phones250 W3 h 3415 h 10
Add a space heater1,500 W0 h 362 h 32
◆ This
A 1 kWh unit runs a fridge for seventeen hours, or a fridge plus lights, internet and phone charging for three and a half. Both are true. The first number is the one on the marketing, and the second is the one you will live.

Freezers

A chest freezer is generally easier than a refrigerator, not harder. It is better insulated, opened less, and its duty cycle is often lower once at temperature — typically 40–60 W average for a domestic chest freezer.

The catch is the starting surge, which is usually larger than a fridge's. Check the unit's surge rating, and be aware that surge claims are not always reliable — one independent test recorded a shutdown at 3,140 W against a 3,900 W claim.[2]

Practical points

Plug in before the outage, not during. Several units here specify UPS switchover between 10 and 30 ms, and the Bluetti AC70's was independently measured at 14 ms.[3] With the fridge permanently plugged into the power station and the power station plugged into the wall, an outage causes no interruption at all.

Do not run it through the wall. Connecting a battery to house circuits needs a transfer switch and, in most US jurisdictions, a licensed electrician. An extension lead to the fridge is a purchase; backfeeding a panel is an installation.

Recharge matters as much as capacity. If power returns briefly, a unit that refills in 54 minutes recovers a full charge where one taking three hours recovers a third.[4]

Keep it charged. A backup at 40% is not a backup.

FAQ

How long will a power station run a refrigerator?

Roughly 17 hours for a 1 kWh unit, 33–36 hours for a 2 kWh unit, and about 72 hours for a 4 kWh unit — assuming a full-size fridge averaging 52 W. All calculated from independently measured usable capacity.

How long does food stay safe without power?

About 4 hours in a closed refrigerator and about 48 hours in a full freezer, per US food-safety guidance. Keeping the door shut is the single most effective thing you can do.

What size power station do I need for my fridge?

Multiply 52 W by the hours you want to cover, then divide by 0.85 to correct for conversion loss. Twelve hours needs about 734 rated watt-hours; twenty-four hours needs about 1,470.

Can a power station run a fridge and other things?

Yes, but for much less time. A 1 kWh unit runs a fridge alone for about 17 hours, or a fridge with lights, router and phones for about 3.5. Size against the full load, not the fridge on its own.

Will it start my fridge's compressor?

Almost certainly — a compressor surge is typically several hundred watts and every unit here has a surge rating well above that. Treat published surge ratings as approximate; at least one has been measured cutting out below its claim.[2]

For the full ranking by measured capacity see the best power stations, or work out your own total with the sizing guide.

Sources

Every figure on this page traces to one of these
  1. OutdoorGearLab. Anker SOLIX C300 Review. Independent lab test · accessed 5 September 2026
  2. APH Networks. Bluetti Elite 200 V2 Review. Independent lab test · accessed 5 September 2026
  3. StorageReview. Bluetti AC70 and AC2A Portable Power Station Review. Independent lab test · accessed 5 September 2026
  4. OutdoorGearLab. EcoFlow Delta 3 Review. Independent lab test · accessed 5 September 2026
SPA
The Solar Power Auditor desk

We do not run our own bench. We collect published laboratory measurements from independent testers, put them beside the manufacturer figures, and show the arithmetic behind every runtime we quote. Where nobody has measured something, we say so rather than estimating it.

How this site works →