AI carbon footprint in EU West - Ireland: emissions per token and per query
Factor version soma-ai-ef-2026.10-dataset-v3.0.0. Last updated . Location-based grid averages, not renewable contracts. Estimates from published benchmarks.
Part of the SOMA AI Emission Factors cluster.
How much CO₂ does an AI query emit when served from EU West - Ireland?
A 1210-token query to a class B model (gpt-4o) served from EU West - Ireland emits about 0.033 g CO₂e, and a million tokens emit 0.0245 kg CO₂e.
Rank 8 of 16 for class B, lowest carbon first (full table). 3.4x the level of EU North - Sweden, Asia Pacific - Singapore is 1.6x higher.
Which grid supplies the electricity in EU West - Ireland?
EU West - Ireland maps to Ireland, a national grid, at 282.5 g CO₂e per kWh for 2023. That is 1.2 times the EU-27 average (236 g) and 0.6 times the global average (483.2 g). The source is Ember yearly electricity data, CO2 intensity 2023, 282.5 g/kWh (lifecycle, IPCC AR5).
Cloud regions on this grid, 2 in all (1 AWS, 1 Azure).
| Provider | Region id | Region name | Location | Mapping confidence | Mapping note |
|---|---|---|---|---|---|
| AWS | eu-west-1 | Europe (Ireland) | Ireland | high | Provider states country only. |
| Azure | northeurope | North Europe | Ireland | high | Provider states country only (Ireland). |
The note for AWS eu-west-1 reads "Provider states country only.".
Which grids have an intensity close to Ireland?
| Difference g | Grid | Code | g CO₂e per kWh | Ratio |
|---|---|---|---|---|
| -4.2 | Fiji | FJI | 278.3 | 0.985 |
| -5.8 | NPCC New England | NEWE | 276.7 | 0.979 |
| 9.3 | Peru | PER | 291.8 | 1.033 |
| 12.6 | Liberia | LBR | 295.1 | 1.045 |
| -14.3 | Netherlands | NLD | 268.2 | 0.949 |
| -16.3 | ASCC Miscellaneous | AKMS | 266.2 | 0.942 |
| -16.3 | Panama | PAN | 266.2 | 0.942 |
| 18.4 | Nicaragua | NIC | 300.9 | 1.065 |
| -20.1 | Armenia | ARM | 262.4 | 0.929 |
| 21.3 | SERC Virginia/Carolina | SRVC | 303.8 | 1.075 |
Emission factors for EU West - Ireland by model class
| Model class | Electricity kg | Hardware kg | Construction kg | Training kg | Total kg CO₂e | g per query | Water L | mL per query |
|---|---|---|---|---|---|---|---|---|
| Class A (small) | 0.00615 | 0.00027 | 0.00014 | 0.00060 | 0.007 | 0.0098 | 0.032 | 0.045 |
| Class B (mid-size) | 0.02014 | 0.00101 | 0.00046 | 0.00293 | 0.025 | 0.033 | 0.106 | 0.15 |
| Class C (frontier) | 0.07499 | 0.00388 | 0.00169 | 0.01284 | 0.093 | 0.13 | 0.395 | 0.55 |
| Model class | Central | Class range low | Class range high | High case | B200 | Development |
|---|---|---|---|---|---|---|
| Class A | 0.007161 | 0.004506 | 0.01028 | 0.01129 | 0.006316 | 0.008386 |
| Class B | 0.02454 | 0.0162 | 0.02948 | 0.03809 | 0.01765 | 0.02746 |
| Class C | 0.09342 | 0.03984 | 0.1371 | 0.1439 | 0.09589 | 0.1023 |
How much does siting matter for a model served from EU West - Ireland?
Electricity is 82.1% of the class B total here (class A 85.9%, class C 80.3%), with hardware at 4.1%, construction 1.9% and training 12%. Electricity carries most of the 0.0245 kg, so the grid is the biggest lever.
The example company (2,323 million class B tokens a year) emits 57 kg CO₂e from EU West - Ireland. Moving to Netherlands would save 2.4 kg (4.2%) or EU West - UK (London) would save 7.8 kg (14%).
| Region | Grid g per kWh | Class A | Class B | Class C |
|---|---|---|---|---|
| Netherlands | 268.2 | 0.725 | 2.37 | 8.84 |
| EU West - UK (London) | 235.6 | 2.37 | 7.77 | 28.9 |
How much water does AI usage consume in EU West - Ireland?
A class B query of 1210 tokens from EU West - Ireland consumes about 0.15 mL of water and a million tokens 0.106 litres, or 0.0815 litres without the water attributed to hydropower.
| Input | Value |
|---|---|
| On-site WUE, L per kWh | 0.02 |
| WUE denominator | IT load |
| WUE basis as published | withdrawal |
| Withdrawal to consumption | 0.6667 |
| EWIF, L per kWh | 1.476 |
| EWIF without hydro, L per kWh | 1.132 |
| AWARE factor | 1.17 |
| Confidence | published-site |
| Source | AWS; AWS Region Europe (Ireland) eu-west-1; 2025 |
| Model class | On site | Grid electricity | Total | Total without hydro | AWARE-weighted on site |
|---|---|---|---|---|---|
| Class A | 0.000242 | 0.0321 | 0.0324 | 0.0249 | 0.000283 |
| Class B | 0.000792 | 0.105 | 0.106 | 0.0815 | 0.000927 |
| Class C | 0.00295 | 0.392 | 0.395 | 0.303 | 0.00345 |
Where do these numbers come from and how do I cite them?
Figures per query are one plain exchange, before the Tier 2b reasoning uplift. Electricity term. GPU energy per token from the ML.ENERGY Leaderboard v3, a facility overhead of 1.20, and grid carbon intensity from EPA eGRID 2023 and Ember 2023. Embodied hardware, data-centre construction and amortised training terms are documented on the methodology page. Water combines on-site cooling water (Li et al. 2025) with the water embedded in grid electricity (Reig et al. 2020). The modelled range of the models inside a class, around the class value, is −41% to +48% for class A, −39% to +24% for class B and −67% to +55% for class C.
Dataset: AI Inference Emission and Resource Factors for Corporate GHG Inventories (SOMA AI Emission Factors (SAEF)), Llopis-Garcia, G., SOMA AI, 19 October 2026, CC BY 4.0, concept DOI 10.5281/zenodo.20443585, which resolves to the latest version (this version 10.5281/zenodo.23371447, earlier version 2 values at 10.5281/zenodo.22767475). Method: Accounting for Purchased AI Inference in Corporate GHG Inventories: A Token-Based Method, arXiv:2606.10660, under peer review.
Partners can fetch any row from the AI factors API, and the whole table is a CSV download.
Cite as "0.0245 kg CO₂e per million tokens, class B, EU West - Ireland, SAEF, 19 October 2026" and link somaai.earth/ai-footprint/factors/eu-ireland.