Running cost, compared

Every row below is two appliances doing a comparable job, costed on the same model at the same rate, with each one’s operating assumption printed on its page. Where the honest answer is “these two cost the same and the marketing says otherwise”, the page says that instead.

Annual running cost at 18.34¢/kWh, the EIA residential average for June 2026. Sorted by the size of the gap.

Every comparison

• Heat pump vs electric furnace

  $1,230 a year apart

  Is a heat pump actually cheaper to run than electric resistance heat?

• Ceiling fan vs air conditioning

  $411 a year apart

  Can a ceiling fan replace air conditioning?

• Heat pump water heater vs a standard electric tank

  $322 a year apart

  How much does a hybrid water heater actually save?

• Chest freezer vs a second fridge in the garage

  $67 a year apart

  Which costs more to run: a chest freezer or an old second refrigerator?

• Heat pump dryer vs a vented electric dryer

  $53 a year apart

  Does a ventless heat pump dryer really use half the electricity?

• Air fryer vs electric oven

  $44 a year apart

  Does an air fryer really use less electricity than the oven?

• Level 2 vs Level 1 EV charging

  $38 a year apart

  Is it cheaper to charge an electric car slowly?

• Oil-filled radiator vs fan heater

  $33 a year apart

  Is an oil-filled radiator cheaper to run than a fan heater?

• Window vs portable air conditioner

  $32 a year apart

  Is a portable air conditioner more expensive to run than a window unit?

• LED vs incandescent

  $17 a year apart

  How much does switching one light bulb actually save?

How to read a comparison

Three things decide every one of these, and only one of them is usually discussed. The first is the efficiency ceiling. Anything that makes heat from electricity — a resistance element in a furnace, a water heater, a dryer, a space heater — converts it at 100%, and that is the ceiling rather than a selling point. Anything that moves heat instead, which means anything with a compressor, delivers two to four units of heat per unit of electricity. No amount of build quality lets the first kind catch the second, which is why the largest gaps on this page are all the same physics.

The second is the duty cycle: the share of its switched-on hours an appliance is actually drawing its rated power. A 15,000 W appliance cycling on a thermostat is not a 15,000 W load, and multiplying the plate rating by the hours — which is how most published comparisons are done — overstates the answer, sometimes by several times. Every figure here uses the duty cycle instead, and every appliance page prints the one it used.

The third is the hours, and it is the one that is yours rather than the appliance’s. The duty cycle is a property of the machine; how long you run it is not. That is why every page here carries your rate and links to the sliders on each appliance’s own page — the comparison is stable, but the size of the gap is personal.

Why some of these gaps are large and others are not

The pairs fall into three kinds, and knowing which one you are looking at tells you most of the answer before you read the number.

• One moves heat, the other makes it. The heat pump comparisons. These produce the widest gaps on the site — up to $1,230 a year — because the two sides are not doing the same physics.
• One does less work than the other. A ceiling fan does not cool a room and an air fryer does not cook a family meal. The gap is real and large, but so is the difference in what you get, and the honest use of these pages is to decide when the smaller job is the one you actually need.
• Same technology, different packaging. A fan heater against an oil-filled radiator, Level 1 against Level 2 charging, a window unit against a portable. Here the gaps are the smallest on the page, and the useful finding is usually that the marketing claims a difference the meter cannot see.

A large ratio and a large amount are also different questions. Swapping one incandescent bulb for an LED cuts its running cost by about six to one and saves $17 a year, which is a striking ratio attached to a trivial sum. It matters because a house has thirty of them, not because one of them does. The reverse also happens: a gap of a few percent on heating is worth more than a bulb ever will be.

What these pages do not do

They do not tell you what anything costs to buy. Where replacing one with the other is a real decision, the page asks for the quote you were given rather than asserting an installed price — quotes for the same job differ by thousands with the house, the region and the incentive, and a made-up figure would be the least defensible number on the page while looking like the most important one.

They also do not model your climate, your insulation, your tariff structure or your household. A heat pump’s advantage narrows in a cold snap; an air conditioner’s hours quadruple between Maine and Arizona. Each page states what would change its answer, in its own words, because a comparison with no caveat is an advertisement. Where a figure matters enough to act on, measure the appliance rather than trusting any table, including this one.

Source: U.S. Energy Information Administration, Electric Power Monthly, Table 5.6.A, June 2026. U.S. Government work, public domain.