Impacts of energy consumption
For any study object, operational impacts are derived from electricity consumption. This page explains the conversion into environmental impacts and how reference electricity factors are constructed. Their variation by country, season and generation mix is a major driver of results.
Convert electricity consumption into operational impact
The conversion follows:
The equation is generic. Index identifies an environmental indicator for which a compatible factor is available. Adapt accordingly: kg CO₂ eq. for GWP, characterized m³ eq. for WU, and so on.
Factor construction
Country electricity factors
The methodology, resources and results are available in the open-source digital4better/data repository.1
The dedicated database aims to:
- cover countries and selected state or regional subdivisions.
- provide annual or monthly factors according to available coverage.
- include all nine environmental indicators.
- use scientific and technical references that can improve on older datasets.
Calculation method
Factors represent electricity consumed within a geographical region. They account for imports and exports, unlike a production-only approach.
For provinces or states, national imports and exports are distributed across subdivisions in proportion to their electricity generation.
Equation
Generation technologies
The calculation uses nine technology categories:
| Family | Technologies | Main characteristic |
|---|---|---|
| Fossil | Coal, gas, other fossil | High GWP and acidification contributions |
| Low-carbon grouping | Wind, solar, bioenergy, hydro, other renewable, nuclear | GWP and other impacts vary by technology and inventory; interpretation depends on the factor used |
Two factors per country
Two factors are available to investigate renewable-electricity scenarios:
- Standard factor
Covering the complete national mix.
- Renewable factor
Covering only generation classified as renewable.
The reference calculation retains the complete average mix. The renewable factor supports a separate sensitivity analysis: contractual REF alone does not reconstruct physical supply or reduce an SCI score. See Software layers and virtualization and Datacenter.
Geographical and temporal variability
The same electricity consumption can produce different impacts depending on geography, period and generation mix.
Geographical example: GWP
| Country | Year | GWP factor, g CO₂e/kWh |
|---|---|---|
| France | 2025 | 31.93 |
| Sweden | 2025 | 12.82 |
| Germany | 2025 | 329.63 |
| Poland | 2025 | 644.15 |
| China | 2025 | 589.30 |
| United States | 2025 | 354.80 |
Values come from the versioned D4B dataset2, accessed on 16 October 2026 and also used by the simulators. Ex: 1,000 kWh corresponds to 31.93 kg CO₂e in France and 644.15 kg in Poland, with identical consumption and electricity boundaries. These are not complete server footprints.
Temporal variability
The default conventions are:
- Annual factors
For scoping assessments.
- Monthly factors
When seasonality materially changes the result, such as seasonal hydropower or summer solar generation.
Temporal example: GWP
Within the same country, the factor also varies from month to month. Values below are in g CO₂e/kWh, from the same D4B snapshot as the geographical examples.3
| Country | January 2025 | July 2025 |
|---|---|---|
| France | 42.66 | 26.36 |
| Germany | 372.63 | 317.31 |
| Brazil | 53.46 | 83.13 |
For 1,000 kWh, electricity impacts in France are 42.66 kg CO₂e in January versus 26.36 kg in July. In Brazil, they rise from 53.46 to 83.13 kg CO₂e. These two months illustrate observed variation, not a seasonal rule that holds every year.
A monthly assessment applies each month’s factor to its corresponding consumption, then sums the impacts. When consumption varies, the average factor is weighted by each month’s kWh. Monthly averages do not describe hourly variations.
Freshness depends on the latest period actually present in the file. An intended monthly update frequency does not prove freshness or completeness. Simulators here use a complete year, 2025, rather than extrapolating available months of 2026.
Interpreting electricity contracts
The ADEME Product Category Rules (PCR) for digital services, version 2 of June 2026, use the geographical consumption mix, including imports. Guarantees of origin and offsets do not reduce that result. They provide specific treatment for on-site generation, with its full lifecycle and conditions on certificates; an off-site renewable contract does not replace the local mix.4
D4B retains the country consumption factor for geographical reporting. The renewable factor supports a separate sensitivity analysis. Its use for the cloud REF share is described in the cloud convention, with its specific data and limitations.
Resources
Electricity technology impacts are primarily based on:
- coal, gas, wind, solar, hydro and nuclear: Life cycle assessment of electricity generation options5, UNECE.
- other fossil, bioenergy and other renewable: New approach for assessing and optimising the environmental performance of multinational electricity sectors: A European case study6, September 2022.
These technology inventories describe Europe. Applying them worldwide where more representative data are unavailable introduces a geographical limitation that must remain visible.
National mixes draw notably on Ember7, with monthly data for covered countries and annual data for broader coverage. Additional sources cover subdivisions, including Canada8 and the United States9.
Assumptions and limitations
- Average and marginal electricity mixes answer different questions. This methodology uses average mixes by default, consistent with its attributional perspective.
- European technology inventories may require extrapolation to other regions.
- Every assessment must identify the version of its D4B factors; see Time and versioning.
- Renewable factors must not be used without reviewing the contractual and physical supply conditions.