How many watts does a refrigerator actually use in a day?
Power basics Updated 2026-08-01
A household refrigerator does not draw its rated power continuously. The compressor cycles, so the number you want is running watts (while the compressor is on) multiplied by the duty cycle (the fraction of the day it is on) multiplied by 24 hours.
The nameplate amperage on the door sticker is not the running power. It is a maximum that includes the defrost heater and the fans, and multiplying it by twenty-four hours overstates a fridge’s daily energy by several times.
The three figures, and how they multiply
Full-size refrigerator
No refrigerator manufacturer we checked publishes a running-watts figure. What they publish is an annual kilowatt-hour number on the energy label and an amperage on the nameplate — so the running figure below is derived from the annual one and the duty cycle, and says so.
Straight off the appliance maker’s published energy label.
See the load record →The share of the day the compressor is actually running, from an appliance maker’s own support documentation. In a hot, humid room it approaches one.
The annual figure spread over the hours the compressor is actually running.
Derived: annual energy divided by the hours in a year, then divided by the duty cycle — so that running watts times duty cycle returns the label figure it came from.
The number to size a battery against. This is what a fridge costs you in a day.
officialverified 2026-08-01full-size refrigerator record
Both directions of that arithmetic have to agree or the record is wrong, and this is where most published figures fall apart. A page that takes an average annual figure and then multiplies it by a duty cycle has applied the duty cycle twice — the annual average already contains it — and comes out roughly a sixth too optimistic on runtime. A page that takes the nameplate amps and multiplies by twenty-four has used a maximum as an average and comes out several times too pessimistic.
Why the door sticker is not the answer
The amperage printed on a refrigerator’s nameplate rates the circuit it is allowed to be plugged into, not the compressor. It has to cover the worst simultaneous case — compressor, condenser fan, evaporator fan and the defrost heater — because that is the figure an electrician needs when deciding what breaker protects the socket. Multiplying it by your mains voltage gives a volt-ampere ceiling several times the running power, and multiplying that by twenty-four hours gives a daily energy figure a household refrigerator has never used in its life. It is the single most common way people conclude they need a machine four times larger than they do.
The defrost heater is worth a sentence of its own, because it is the one part of the sticker that is real and intermittent. It draws hundreds of watts for a short period on a cycle measured in hours or days, and it is already inside the annual energy figure on the label — which is why the annual number is the honest starting point and the nameplate is not. It does mean the instantaneous draw of a fridge is spikier than a single running-watts figure suggests, and that is a reason to keep headroom in the inverter, not a reason to size the battery differently.
What the same treatment does to other cold loads
Watt-hours per day, three cold appliances
All three cycle. None of them draws its running watts for twenty-four hours, and the differences between them are much larger than brand-to-brand differences within a category.
Smaller box, better insulation, no door-opening allowance in the test — it beats the refrigerator despite running colder.
See the record →A variable-speed compressor that modulates instead of cycling hard, on a European energy label.
See the record →officialverified 2026-08-01appliance records
Where the duty cycle lands in a runtime
When our engine runs BLUETTI Elite 200 V2 against this refrigerator it does not divide by the running watts. It divides by 47 W — the running figure weighted by the duty cycle — plus 10 W for the machine’s own consumption, and returns 28 h. Use the running watts there instead and the same pair comes out roughly a fifth short.