Picture two T-shirts on a rail. Same cotton, same weight, same cut, same colour. Nobody could tell them apart. Their carbon footprints can still differ — and much of the reason lies in how they were made, which is nothing you could see.
A shirt is made of energy as much as cotton
Most of a garment’s footprint is created in the mill: spinning yarn, knitting or weaving fabric, dyeing and finishing it. Looking beyond shipping in a product’s carbon footprint walks through those stages. What they share is energy. A study of cotton shirts made in China found energy use — electricity above all — to be the main driver of the footprint.
So the question is not only how much energy a factory uses. It is where that energy comes from: the electricity it draws from the grid, and the heat it raises on site.
Grid electricity: not every kilowatt-hour is equal
The carbon behind a unit of electricity depends on how a country generates it. Grids that lean on coal put a lot of CO₂ behind every kilowatt-hour. Grids built on hydro, wind, solar or nuclear put very little.
Ember’s electricity data, published through Our World in Data, shows how wide that gap is.
Country averages: carbon intensity of electricity generation, 2025. Grams of CO₂e per kilowatt-hour generated. Source: Ember (2026), via Our World in Data.
| Country | g CO₂e per kWh (national average) |
|---|---|
| Bangladesh | 696 |
| India | 670 |
| China | 525 |
| Türkiye | 475 |
| Vietnam | 461 |
| Italy | 285 |
| Portugal | 128 |
| Ethiopia | 23 |
| World average | 458 |
Country-level information provides context. A particular product’s footprint depends on its own production conditions and lifecycle.
Read as averages, the numbers still show the range. On these figures, an average kilowatt-hour from Bangladesh’s grid carried more than five times the CO₂e of one from Portugal’s, and around thirty times one from Ethiopia’s, where most power comes from hydro. A factory with its own solar panels or a supply contract for clean power can sit well away from its country’s average.
Process heat: not only the electricity
Grid electricity is only part of the picture in a textile mill. Dyeing and finishing — the largest single stage in the apparel industry’s emissions, according to the Measuring Fashion study — need hot water and steam, and that heat often comes from a boiler on site. What the boiler burns makes a difference of its own. Two mills in the same country, on the same grid, can still turn out shirts with different footprints.
Why an average misses this
Many carbon estimates for clothing start from averages: a typical cotton T-shirt, a typical factory, a typical country. Averages are useful for understanding an industry. They cannot tell two shirts on the same rail apart, because the things that separate them — which mill, which power, which boiler — are exactly what an average smooths away.
That gap is also an opportunity. A brand that moves its dyeing to a mill on cleaner power, or helps a supplier replace an old boiler, can lower the carbon of the shirts that come off that line.
Grids are changing
None of these numbers are fixed. Many countries are adding wind and solar, and a factory’s footprint can fall as its grid changes, without a single change inside its walls.
When credible production improvements become visible in everyday decisions, they can become commercially relevant. Karbonomics is developing a system to help make that connection.