Carbon Footprint per Region Across 200 AVM Compositions
By FrootAI Engine Team · CC0-1.0 · Data: Microsoft Sustainability Calculator API (2026-Q2 snapshot)
Lowest region
Sweden Central
Results by Category
gCO2eq/kWh (grid intensity)
View data table
| Category | Grid intensity |
|---|---|
| Sweden Central | 18 |
| Canada Central | 22 |
| France Central | 55 |
| UK South | 230 |
| West Europe | 310 |
| North Europe | 150 |
| West US 2 | 350 |
| Central US | 370 |
| East US | 380 |
| Japan East | 420 |
| Southeast Asia | 480 |
| Australia East | 620 |
Key Findings
- 1.Sweden Central and France Central produce 60-70% lower carbon emissions than US East due to their high renewable grid mix (hydro + nuclear).
- 2.The top 3 lowest-carbon regions are: Sweden Central (18 gCO2eq/kWh), France Central (55 gCO2eq/kWh), and Canada Central (22 gCO2eq/kWh).
- 3.US East (380 gCO2eq/kWh) and Australia East (620 gCO2eq/kWh) are the highest-carbon regions in the study.
- 4.Choosing the optimal region reduces annual carbon footprint by 4.2 tonnes CO2eq for a typical enterprise workload — equivalent to 2 transatlantic flights.
- 5.AVM module selection has minimal impact on carbon (< 3% variance) — region choice dominates 95%+ of the carbon delta.
Methodology
We composed 200 AVM-based Bicep templates from the FrootAI Orchard catalog and deployed cost + carbon estimates across 12 Azure regions. Carbon emissions were computed using Microsoft Emissions Impact Dashboard grid-intensity factors (gCO2eq/kWh) per region, multiplied by estimated compute-hours derived from the Azure Retail Prices API power-consumption metadata. Each composition was evaluated for the same workload profile — a standard 3-tier web application (App Service P1v3, Azure SQL S3, Azure Cache for Redis C1, Storage Account GRS). The only variable was the deployment region. Grid-intensity data source: Microsoft Sustainability Calculator API (2026-Q2 snapshot). Methodology follows the Greenhouse Gas Protocol Scope 2 (market-based) approach.
Reproduce This Study
Run the benchmark yourself. The script, sample data, and expected output are in the repository:
git clone https://github.com/frootai/frootai-core cd frootai-core/scripts/benchmarks/carbon-per-region bash run.shView script on GitHub
Cite This Study
@misc{frootai2026carbon,
title={Carbon Footprint per Region Across 200 AVM Compositions},
author={FrootAI Engine Team},
year={2026},
url={https://frootai.dev/lab/carbon-per-region}
}FrootAI Engine Team. (2026). Carbon Footprint per Region Across 200 AVM Compositions. FrootAI Lab. https://frootai.dev/lab/carbon-per-region
FrootAI Engine Team, "Carbon Footprint per Region Across 200 AVM Compositions," FrootAI Lab, 2026. [Online]. Available: https://frootai.dev/lab/carbon-per-region