Carbon intensity of electricity by location, offline, anywhere
Source:R/compute_footprint.R
co2int.RdThe grams of CO2-equivalent emitted per kilowatt-hour of electricity at a
location, with the year and source of the figure. The table is bundled with
the package, so it works offline in any country: every country at the latest
year published by Our World in Data from Ember's yearly electricity data (the
latest year each grid reports: 2025 or 2024 for nearly all, older for a few
grids, Ukraine (2022) among them, named in year), sub-national zones
from Electricity Maps' 2024 yearly data as redistributed in the Green
Algorithms data release v3.1 (Canadian provinces and territories, US
balancing authorities, Australian states, Indian and Japanese regions,
Brazilian and Chilean systems, ...), and the world average. All figures are
lifecycle intensities (generation plus upstream), the basis both sources
publish.
Value
co2int(): a list with g_per_kwh,
location, name, year, basis, source
and note (a character vector, empty unless a fallback or detection
was involved). co2inttab(): the full table as a data
frame with the same columns plus iso3.
Details
A location is matched by its code ("CA", "CA-ON",
"US-CAL-CISO"), its ISO 3166-1 alpha-3 code ("CAN") or its
name ("Canada"), case-insensitively. An unknown sub-national code
falls back to its country and says so in note. NULL detects
the location with dtctloc.
Provenance and licence of the table
The country rows are Ember's yearly electricity data as published by Our
World in Data (CC BY 4.0); the sub-national zones and the Canadian
provinces come from the Electricity Maps yearly files redistributed in
the Green Algorithms data repository (v3.1), under that repository's
terms, and the default PUE from the same repository. Every row names its
source and year. Values are reproduced unchanged: Ember reports three
small diesel grids (Montserrat, Saint Helena, Uzbekistan) at exactly
1000 gCO2e/kWh and Turkmenistan's gas fleet at 1306, and the table does
not round or cap them. Kosovo appears once, under the code XK.
References
Ember (2026). Yearly Electricity Data, via Our World in Data, ourworldindata.org/grapher/carbon-intensity-electricity. Electricity Maps (2025). 2024 Yearly Carbon Intensity Data, via Green Algorithms data v3.1. Lannelongue, L., Grealey, J. and Inouye, M. (2021). Green Algorithms: quantifying the carbon footprint of computation. Advanced Science 8(12), 2100707.
Examples
co2int("CA-ON")
#> ╭── Carbon intensity of electricity ──────────────────────────────────────────╮
#> │ Intensity 90.97 gCO2e/kWh │
#> │ Location CA-ON (Canada, Ontario) │
#> │ Year 2024 │
#> │ Basis lifecycle │
#> │ Source Electricity Maps (2025). Canada 2024 Yearly Carbon Intensity │
#> │ Data (Version January 27, 2025). Electricity Maps (via Green │
#> │ Algorithms data v3.1) │
#> │ │
#> │ The emissions of producing one kilowatt-hour of electricity on this │
#> │ grid; co2fp() multiplies a computation's energy by it. │
#> ╰─────────────────────────────────────────────────────────────────────────────╯
co2int("Canada")$year
#> [1] 2025
co2int("CA-XX")$note
#> [1] "no row for zone CA-XX; its country CA used"
co2int(120)$source
#> [1] "user-supplied carbon intensity"
nrow(co2inttab())
#> [1] 378
# the Canadian provinces in the table, lowest intensity first
tab <- co2inttab()
ca <- tab[
startsWith(tab$location, "CA-"),
c("location", "name", "g_per_kwh", "year")
]
head(ca[order(ca$g_per_kwh), ])
#> location name g_per_kwh year
#> 236 CA-NL Canada, Newfoundland and Labrador 27.60 2024
#> 227 CA-QC Canada, Quebec 35.04 2024
#> 322 CA-MB Canada, Manitoba 52.89 2024
#> 367 CA-BC Canada, British Columbia 70.34 2024
#> 241 CA-PE Canada, Prince Edward Island 70.94 2024
#> 313 CA-ON Canada, Ontario 90.97 2024
# a location the table does not know is an error, not a silent world average
try(co2int("Atlantis"))
#> Error : Unknown location 'Atlantis'. Use a code, alpha-3 code or name from co2inttab(), or pass a number in gCO2e/kWh.