Methodology
Every number on this site is either published by the US Energy Information Administration or derived from it by arithmetic we show you. This page explains exactly how, and where the method is weak.
The calculation
There is no proprietary model here. Running cost is a four-term multiplication, and we would rather show it than dress it up:
watts ÷ 1000 × hours per day × duty cycle = kWh per day
kWh per day × rate in cents ÷ 100 = dollars per day
Monthly figures multiply the daily figure by 30.44, the average length of a calendar month. Annual figures multiply by 365. We do not compound, discount or apply seasonal weighting to the headline numbers, because doing so would bury the arithmetic that lets you check our work.
Where the electricity rates come from
State rates are the average retail price of electricity to residential customers, published monthly by the US Energy Information Administration in Electric Power Monthly, Table 5.6.A. We pull them through the EIA's public API and store the most recent twelve months so you can see the trend rather than a single snapshot. The figures currently shown are for May 2026.
This is a genuine, independently published measurement, and it is the strongest number on the site. It is also a state average, which is the source of most of the difference between our figure and your bill. More on that below.
Where the appliance wattages come from
Wattages are compiled from manufacturer nameplate ratings and published appliance energy-use references, then sanity-checked against the typical range for that class of appliance. Each appliance carries a low and high figure as well as a typical one, and the range is shown on every page precisely because the typical value will not match every model.
Being straight about this: the wattage, the default hours of use and the duty cycle are estimates, not measurements. We have not put a meter on every appliance in this database. They are reasonable central values for a class of product, and they are the least certain numbers on the site. That is exactly why every page puts an editable calculator above the prose - your nameplate rating and your actual hours beat our defaults every time.
Duty cycle, and why it matters more than people expect
Many appliances do not draw their rated power continuously. A refrigerator compressor cycles; a thermostat-controlled heater switches off once it reaches temperature. Treating a 1,500W space heater as though it draws 1,500W for every hour it is switched on can overstate its cost by a factor of two or more.
Where an appliance cycles, we apply a duty cycle - the share of "on" time it actually draws power - and we show it on the page rather than hiding it inside the result. Duty cycles are the second-least certain input here, because they depend heavily on your home's insulation, the ambient temperature and the thermostat setting.
Seasonal and regional context
Each state page carries a classification of the local climate and electricity market structure. These drive the contextual notes on cost pages: whether an appliance's load falls in winter or summer in that particular state, and whether the published state average is a good guide to what you personally pay. In deregulated states with retail supplier choice, it often is not.
Known limitations
- State averages hide real variation. New York is the clearest example: upstate hydroelectric power is cheap while downstate delivery charges are among the highest in the country, and a single state number conceals that gap entirely.
- We do not model tariffs. Time-of-use pricing, tiered rates, demand charges, fixed monthly service fees and local taxes all affect your real bill and none of them appear in a per-kWh average.
- Wattages are class typicals. A ten-year-old appliance and a current efficient model can differ substantially even under the same product name.
- We have not independently measured these appliances. If we ever publish measured readings, they will be labelled as such and dated.
- Rates move. Figures update monthly when the EIA publishes. A page read months after its last update reflects the rate at the top of the page, not today's.
Corrections
If a figure here looks wrong, we would genuinely like to know, particularly if you have a meter reading that contradicts one of our estimates. Write to contact@costtorunit.com. We correct errors rather than quietly deleting pages, and material corrections are noted on the page itself.
Currently tracking 40 appliances across 51 states and territories. Rate data last published for May 2026. Content last reviewed 2026-07-26.