How CoolingMetrics Calculates Running Cost (Our Methodology)

CoolingMetrics ranks every air conditioner by one number: cost per hour to run. Here’s exactly how that number is produced — no black box.

The core formula

Cost per hour = rated power (kW) × electricity rate (per kWh)

Rated power is how many watts the unit draws while cooling. Divide by 1,000 for kilowatts, multiply by your electricity price. The monthly estimate is simply cost per hour × hours per day × 30.

Where the wattage comes from

Listings rarely state power draw directly, so we resolve it in priority order:

1. Stated watts — taken straight from the title or feature bullets when present (most accurate, but uncommon).
2. BTU ÷ stated EER — when the listing gives a full-load EER, we divide cooling capacity by it.
3. BTU ÷ EER estimated from SEER — when only a seasonal SEER/SEER2 is quoted, we first convert it to an estimated EER (see below), then divide. The SEER itself is still shown in the comparison, labelled as SEER.
4. BTU ÷ typical EER for the type — otherwise we apply a representative efficiency for that AC type.

Anything in categories 3 and 4 (and some of category 2) carries the est. marker in the comparison, so you always know when a figure is derived rather than measured.

Why we don’t use SEER directly

EER measures efficiency at full load; SEER (and SEER2) is a seasonal average that includes gentle part-load running, so it is always a higher number. Dividing BTU by a SEER would therefore understate the watts a unit actually pulls when it’s working hard — and understate its running cost. When a listing quotes only SEER, we convert it to an approximate full-load EER (using the U.S. Department of Energy relationship EER ≈ -0.02 × SEER² + 1.12 × SEER) before working out power, and mark the result estimated. The original SEER is still displayed so you can see it. For the full explainer see SEER vs EER explained.

Typical efficiency assumptions (when none is stated)

Portables sit lowest because their single-hose designs are inherently less efficient; ductless mini-splits sit highest. These are deliberately conservative type averages — a unit’s stated EER (or an EER estimated from its SEER) always overrides them when available.

Electricity rates by market

Each country uses an approximate average domestic rate in its local currency as the default:

Rates move, and yours may differ — so the comparison has a “Your rate” box. Enter your actual tariff and every cost-per-hour and monthly figure recalculates instantly.

What we filter and why

To keep the ranking honest and useful:

• Evaporative “air coolers” are hidden by default — they’re not refrigerant air conditioners and don’t compare like-for-like (toggle them back on if you want).
• 12V/24V camping and automotive units are filtered by the “Home use only” default, so tiny outdoor units don’t dominate the cheapest-to-run list.
• Near-duplicate listings (the same product from multiple sellers) are collapsed to one row, keeping the best by your current sort.

Data and freshness

Product data and prices come from Amazon (UK & US, plus major EU stores) and are refreshed regularly; the comparison shows when prices were last updated. Where Amazon doesn’t publish a spec, the field is left blank rather than guessed (except power, which is estimated and flagged as above).

No thumb on the scale

The default order is always lowest cost-per-hour, and you control every sort and filter. Affiliate links, where present, never influence the ranking or which products show — see our affiliate disclosure.

Next: How much does it cost to run an air conditioner? · SEER vs EER explained · air conditioners ranked by cost to run.