Grid Carbon Intensity by Country: Why One Number Never Fits All

Last reviewed: 2026-07-30

The same unit, a 25-fold spread

A kilogram of CO2e per kilowatt-hour is a fixed unit, but the number attached to it depends entirely on whose grid generated the kilowatt-hour. The table below lines up national grid emission factors, each with its own year and source, because a "grid carbon intensity" figure quoted without a date and a country is close to meaningless: these numbers move year to year as generation mixes shift, and they vary by generation source, not just by country.

Countrykg CO2e/kWhYearDominant sourceSource
Norway0.0282024Hydro (~90%+)Ember, Global Electricity Review 2025
France0.0412024Nuclear (~65%)Ember, Global Electricity Review 2025
Sweden0.0352024Hydro + nuclearEmber, Global Electricity Review 2025
Brazil0.1102024Hydro (majority)Ember, Global Electricity Review 2025
United Kingdom0.1772025 factorsGas + windUK DESNZ conversion factors 2025
Germany0.3302024Coal + gas + wind, mixedEmber, Global Electricity Review 2025
United States0.3732022Gas + coal, mixedEPA eGRID2022 (823.1 lb CO2/MWh)
China0.5262024Coal (majority)Ember, Global Electricity Review 2025
Australia0.5252024Coal (majority)Ember, Global Electricity Review 2025
Poland0.5892024Coal (majority)Ember, Global Electricity Review 2025
India0.6702024Coal (majority)Ember, Global Electricity Review 2025
South Africa0.6992024Coal (~80%+)Ember, Global Electricity Review 2025

Norway to South Africa is roughly a 25-to-1 range on the same unit. That is not measurement noise; it reflects genuinely different physical infrastructure. A kWh drawn in Oslo and a kWh drawn in Johannesburg are the same amount of energy delivered to a wall socket, and utterly different amounts of carbon released to deliver it.

Why the spread is this large

Hydro and nuclear generation both produce electricity with near-zero direct CO2 emissions per kWh generated (their lifecycle footprint, from construction and fuel mining, is small but nonzero and not counted in the operational figures above). Norway, France, Sweden and Brazil sit near the bottom of the table because each built out one or both of those sources at national scale decades ago, for reasons that had nothing to do with climate policy at the time — Norway for hydro terrain, France for energy independence after the 1973 oil shock. Coal, by contrast, emits roughly twice the CO2 per kWh that gas does for the same electricity output, and countries at the top of the table — South Africa, India, Poland, China, Australia — still generate a majority of their electricity from coal, either because it is locally abundant and cheap or because their grids were built around it before cheaper alternatives existed. The gap between the top and bottom of this table is, in effect, a gap in national industrial history, not a gap in unit definitions.

An annual figure hides an hourly one

Every number in the table above is a generation-weighted annual average, and it hides real variation within each year. A grid with substantial solar capacity is measurably cleaner at 1pm on a sunny day than at 7pm on a windless winter evening, when gas or coal plants are brought online to cover the gap; in a grid like Germany's or California's, the carbon intensity of a kWh can more than double between those two moments. The annual average in the table is the right number for comparing countries against each other or for a yearly carbon disclosure; it is the wrong number for deciding when, within a day, to run a dishwasher or charge an EV for the least carbon-intensive kWh, which is a question several grid operators now answer with real-time, not annual, data.

Where this site's own 0.445 figure fits

The carbon-conversion tools on this site use 0.445 kg CO2e/kWh, the IEA's published 2024 global average power-sector carbon intensity (IEA, Electricity 2025), documented in the CO2e guide. That figure does not appear as a row in the table above because it is not a national figure at all: it is a generation-weighted average across every grid on Earth, including all twelve countries above and every one not listed. It sits inside this table's range, closer to the mixed-fuel middle (Germany, the US) than to either extreme, precisely because the world's electricity is generated by a mix of very clean and very dirty grids, and the large coal-heavy generators (China alone accounts for roughly a third of global generation) pull the average up from where Europe and the Americas alone would place it, while gas-heavy and low-carbon grids elsewhere pull it back down.

Applying 0.445 to a Norwegian kWh overstates its footprint by more than an order of magnitude; applying it to a South African kWh understates it by more than 35%. A single global constant is defensible as a default when a user's grid is unknown, and indefensible as a substitute for a real national or regional figure once one is available — which is exactly the caveat the Methodology page already states about this figure, and this table exists to make concrete with real countries and real numbers.

See also: Methodology and Editorial Policy -- how conversion factors on this site are sourced and reviewed.