What "kg CO2e" Actually Means

A unit built to make different gases comparable

Carbon dioxide equivalent, CO2e, is not a measurement of carbon dioxide. It is a common currency for comparing the warming effect of any greenhouse gas — methane, nitrous oxide, various refrigerant gases — against the effect of an equal mass of CO2 itself. When a figure is expressed as "10 kg CO2e," it means: over the stated time horizon, this quantity of gas, whatever it actually is, traps as much additional heat as 10 kg of pure CO2 would.

GWP100: the exchange rate between gases

The exchange rate between a gas and CO2e is its Global Warming Potential, almost always quoted over a 100-year horizon (GWP100), the IPCC's standard convention, though 20-year GWP figures exist too and give short-lived gases like methane a much larger number, because methane's warming effect is concentrated in its first two decades in the atmosphere rather than spread evenly over a century. GWP100 for CO2 itself is 1 by definition; every other gas is expressed relative to it.

Worked example

Take this site's own grid factor and a real consumption figure: the EIA's most recent published average, 10,791 kWh a year for a US residential electricity customer (2022). At 0.445 kg CO2e per kWh, that is 10,791 × 0.445 = 4,801.995 kg CO2e, or roughly 4.8 tonnes CO2e, from electricity alone in a year. That figure inherits every caveat below: it uses a single global average grid intensity, so a household on a hydro-heavy grid would owe a small fraction of it, and a household on a coal-heavy grid could owe several times as much, and it covers Scope 2 only, purchased electricity, not gas heating, a car, or anything else in the household's footprint.

Why methane's number keeps moving

The number moves for a genuine scientific reason, not because earlier reports were sloppy: GWP depends on how long a gas persists in the atmosphere and how strongly it absorbs infrared radiation, both refined as atmospheric chemistry modelling improves. AR6 also introduced a fossil/non-fossil split for methane for the first time (IPCC AR6 WG1, Table 7.15: 27.0 for non-fossil, 29.8 for fossil), because fossil methane leaves behind a small additional CO2 residue when it eventually oxidises that biogenic methane does not, giving fossil methane a slightly higher effective GWP. Any CO2e figure is therefore implicitly tied to the assessment report it was calculated under; a rigorous carbon disclosure states which one.

This site's own figure, and what it does and does not cover

The kWh-to-CO2e conversion on this site uses 0.445 kg CO2e per kWh, the IEA's published 2024 global average power-sector carbon intensity (IEA, Electricity 2025). That figure is a single global average for one specific year, not a physical constant, and it sits inside a range of real national grids spanning roughly forty to one:

• Grids dominated by hydro or nuclear generation (Norway, France, Sweden) run well under 0.1 kg CO2e/kWh.
• Coal-heavy grids run well above 0.6 kg CO2e/kWh.
• Most large economies with a mixed generation fleet fall somewhere between 0.2 and 0.4 kg CO2e/kWh, below the global average because that average is pulled up by the largest coal-dependent grids.

Applying the global 0.445 figure to a Norwegian household's electricity use would overstate its emissions by roughly an order of magnitude; applying it to a coal-heavy grid would understate them. The Methodology page states this explicitly rather than presenting 0.445 as a universal constant, and explains how this site treats measured, time-varying figures differently from fixed unit definitions.

The other axis: it varies by hour, not just by country

A national average also hides real hour-to-hour variation. Grids with significant solar and wind capacity have a materially cleaner mix at midday, during peak solar output, than at a windless evening peak covered by gas or coal plants brought online to meet demand. In grids with heavy renewable penetration, the carbon intensity of electricity can more than double between a sunny, windy afternoon and a still winter evening. Anyone optimising for genuine impact, timing an EV charge or a dishwasher cycle, is optimising against that hourly figure, which several grid operators now publish in near real time, not against the annual national average this site, and most carbon calculators, necessarily uses.

Why 100 years, and not 20 or 500

The 100-year horizon is a policy convention, not a scientific inevitability, and the IPCC has said so explicitly since its earliest reports: a shorter horizon (20 years) weights short-lived, fast-acting gases like methane much more heavily, which matters if the policy goal is slowing near-term warming; a longer horizon understates their near-term impact while more fairly representing gases like CO2 and nitrous oxide that persist for centuries. GWP100 became the default largely because it was the figure written into the original Kyoto Protocol accounting rules in the 1990s, and reporting frameworks have kept it for comparability ever since, not because 100 years is the uniquely correct choice. A separate metric, GWP*, has been proposed by climate scientists specifically to better capture the different behaviour of short-lived gases like methane against long-lived ones like CO2; it appears in recent IPCC discussion as an alternative under active research, not as a replacement for GWP100 in standard reporting.

What CO2e figures do not include

Grid emissions factors like the one used here cover what corporate reporting calls Scope 2: emissions from purchased electricity generation. They exclude Scope 1, direct on-site combustion such as a gas boiler or a company vehicle, and Scope 3, everything in a supply chain from manufacturing a solar panel to transporting coal. A "carbon footprint" quoted from an electricity-only CO2e figure is a Scope 2 estimate, not a complete accounting: useful for comparing electricity choices against each other, not for a full footprint.

A note on precision

Because grid intensity, GWP values and even which assessment report is current all change over time, a kg CO2e figure is only as trustworthy as its stated vintage. A figure with no year attached to it should be treated as undated and therefore unverifiable, not as more precise for omitting the detail. See the Editorial Policy for how this site reviews and updates figures like this one when a source publishes a newer number.