Key takeaways
- Operational emissions from electricity = electricity use (kWh) × carbon intensity (gCO₂e/kWh).
- Average intensity describes the whole generation mix; marginal intensity describes the generation that changes when demand changes, and short-run and long-run marginal rates can differ a lot.
- Lifecycle factors include building the plant, fuel supply and upstream methane; direct factors count only what the power station emits while running.
- Production-based figures describe electricity generated in a country; consumption-based figures also trace imports and exports.
- National annual averages hide regional and hour-by-hour variation.
What the number means
Carbon intensity divides the emissions from generating electricity by the electricity generated, so it depends on how much of the mix comes from fossil fuels. Ember, whose country data Our World in Data republishes, reports it in grams of CO₂-equivalent per kWh, covering all greenhouse gases converted to CO₂-equivalent over a 100-year horizon.
For a building, operational emissions from electricity are the kWh consumed multiplied by the intensity. For a heat pump, the emissions per kWh of heat delivered are roughly the grid intensity divided by its seasonal coefficient of performance, which is why the same heat pump has very different emissions on different grids.
Average versus marginal emissions
Average intensity spreads the emissions of the whole generation mix over all the electricity produced or consumed. Marginal intensity asks a different question: when demand changes, which generation changes with it, and what does that generation emit?
Short-run marginal rates treat the grid as fixed; long-run marginal rates also include the power plants that new demand causes to be built or retired. For long-lived changes such as electrification, Gagnon and colleagues (2022) argue that short-run marginal rates omit the new generation that added load induces: for US vehicle-electrification targets they estimated 286–336 kg/MWh with long-run marginal rates against 591 kg/MWh with a short-run rate, because in the US new load often leads to more wind and solar being built. Average, location-based figures are the usual basis for emissions reporting; marginal figures are useful for decisions about adding or shifting load.
Lifecycle versus direct emissions
Direct, or operational, factors count only what a power plant emits while running, mostly from burning fuel. Lifecycle factors add the emissions from building the plant, extracting and transporting fuel and disposing of the plant at the end of its life, so every source, including wind, solar, hydro and nuclear, has a value above zero.
Ember's figures aim to cover the full lifecycle, including upstream methane, supply chain and manufacturing emissions. For several sources it uses the IPCC's midpoint lifecycle values, for example 48 gCO₂e/kWh for solar and 24 gCO₂e/kWh for hydro. Ember notes that methane emissions from gas and coal supply are highly uncertain. Always check which basis a figure uses before comparing it with another source.
Where the data comes from
- Ember: free yearly electricity data for 215 countries from 2000, with generation by source and lifecycle emissions, under a CC BY 4.0 licence.
- Our World in Data: republishes Ember's lifecycle carbon intensity of electricity generation by country, with a time series from 1990, as charts and downloads.
- Electricity Maps: grid-zone data at 5-minute, 15-minute and hourly resolution, consumption-based (it traces flows between interconnected grids), with both lifecycle and direct emission factors.
- IEA: CO₂ emission factors from electricity and heat generation for countries worldwide since 1990, sold under licence, with a separate IEA database of life-cycle upstream emission factors for electricity (150 countries from 2015).
- US EPA eGRID: generation and emissions data for most US power plants, which some providers use for regional factors.
Why it matters for electrification and PV
The lower the grid's intensity, the lower the operational emissions of an all-electric building and the stronger the carbon case for replacing fuel boilers with heat pumps. On a high-intensity grid, every kWh saved avoids more emissions.
For on-site PV, the carbon benefit of each kWh depends on the grid electricity it displaces. The national average is the simple, common reference; hourly or marginal data give a better picture when the timing of generation matters. PV's own lifecycle footprint, 48 gCO₂e/kWh in the IPCC midpoint Ember uses, is small compared with the 700 gCO₂e/kWh Ember applies to its 'other fossil' category. Intensities change from year to year as grids add renewables or retire plants, so for buildings expected to last decades, look at the trend as well as the latest year.
How C4B shows grid carbon intensity
C4B shows the carbon intensity of the national electricity grid for the location you search, benchmarks it against other countries and shows the generation mix by fuel, using yearly country data from Our World in Data and Ember. The figures are national, not regional, so they do not capture differences between grid regions within a country.
Frequently asked questions
What units is grid carbon intensity measured in?
Usually grams of CO₂ or CO₂-equivalent per kilowatt-hour (gCO₂/kWh or gCO₂e/kWh). CO₂-equivalent figures include other greenhouse gases, such as methane, converted to the warming effect of CO₂.
What is the difference between average and marginal emissions?
Average emissions spread the emissions of the whole generation mix over all electricity. Marginal emissions describe the generation that responds to a change in demand. Short-run marginal rates assume a fixed grid; long-run marginal rates include the new plants that demand causes to be built.
Should I use lifecycle or direct emission factors?
Lifecycle factors give a fuller picture, including plant construction, fuel supply and upstream methane, and make low-carbon sources non-zero. Direct factors count only emissions at the power station. Use whichever basis your standard or reporting framework requires, and do not mix the two in one comparison.
Why do sources give different figures for the same country?
They can differ in basis (lifecycle or direct), in accounting (production or consumption, which includes imports), in the year covered and in their assumptions for methane and fuel quality.
Does grid carbon intensity change during the day?
Yes, as the mix of generation changes with demand, sun and wind. Services such as Electricity Maps publish data down to 5-minute resolution; annual national averages smooth this variation out.
See it for your site
Grid carbon intensity
The carbon intensity and generation mix of the national electricity grid, compared with other countries.
Preliminary energy analysis
An hourly preliminary energy balance and orientation and aspect-ratio optimization for early design.
Solar radiation and PV potential
Direct, diffuse and global solar radiation, yearly PV potential, optimal panel orientation and geothermal ground temperatures.
See your site's climate before design begins
Pin any location and get climate, comfort and preliminary energy insight your whole team can read. Start with a 7-day free trial, or book a 30-minute demo.
References
- Our World in Data: Lifecycle carbon intensity of electricity
- Ember: Electricity data methodology
- Electricity Maps: Methodology
- IEA: Emissions Factors 2025
- IEA: Life Cycle Upstream Emissions Factors 2024
- Gagnon et al. (2022), Short-run marginal emission rates omit important impacts of electric-sector interventions, PNAS
Related guides
GHI, DNI and DHI
What GHI, DNI and DHI mean, how they relate, their units and data sources, and how they feed façade, shading and PV design.
Heating and cooling degree days
What HDD and CDD measure, how base temperatures work, how to calculate them and how they are used in codes and energy analysis.
SSP climate scenarios
What the five main SSP scenarios mean, how much warming each implies over a building's life and how to choose scenarios for design.