Energy and environmental impacts of rapid data-center expansion in the United Arab Emirates
Abstract
Rapid growth in artificial intelligence (AI) is driving major data-center expansion globally, increasing demand on electric power and water-supply systems. Yet the energy and environmental implications of this growth remain comparatively understudied in hot, water-scarce regions such as the United Arab Emirates (UAE), where announced developments coincide with high cooling requirements, desalination-dependent water supplies, and a natural-gas-dominated electricity system. This study quantifies the electricity, water, and greenhouse gas (GHG) implications of UAE data-center expansion using a bottom-up capacity inventory, modeled operating loads, and a thermodynamic cooling model coupled to national electricity and water supplies. Scenario-aware Monte Carlo simulation propagates uncertainty across 2030–2040 operating scenarios with varying capacity expansion and clean-electricity generation levels. Median annual site electricity demand varies 6.62–37.8 TWh, reaching 11.5% of national generation. Attributed GHG emissions total 3.01–17.6 Mt CO2e annually under baseline grid conditions but fall 12–32% under alternative grids with increased clean generation. Expansion also reduces national clean-generation shares, requiring substantial additional solar photovoltaic capacity to meet UAE decarbonization objectives under some grid scenarios. When clean generation does not expand further to supply the added demand, increased natural-gas generation raises system-wide GHG emissions by 2.27–19.9 Mt CO2e (up to 22%) relative to a no-expansion reference. Water requirements remain high despite modeled power usage effectiveness near 1.22 for AI facilities. Median annual water demand is 10.6–63.1 Mm3, with desalinated cooling make-up comprising over 99% of demand. Dry-cooling configurations can reduce water demand by 98.2% but increase electricity demand by 11–13%, with the net electricity penalty roughly 20 times the desalination electricity avoided. These findings call for expediting clean-electricity deployment and planning water supplies alongside data-center expansion, given that low facility electricity overheads and lower attributed emissions from matching existing clean generation can coexist with substantial system-wide emissions and water requirements.
Related
- Elevated environmental humidity modulates aquaporin expression and impacts leaf hydraulic efficiency in silver birch (Betula pendula)
- Environmental impacts of global industrial and smallholder oil palm plantations: historical dynamics and future climate risks
- Tropical belt expansion and its impacts on climate variability in the Mediterranean from 1980 to 2022
- Environmental risk assessment of wild ungulates: a review of risks, impacts and management
- Marine fuel choices and related life-cycle environmental impacts under global shipping policies
BibTeX
@misc{thaneya2026energy,
title = {{Energy and environmental impacts of rapid data-center expansion in the United Arab Emirates}},
author = {Ahmad Bin Thaneya},
howpublished = {EarthArXiv},
year = {2026},
month = oct,
doi = {10.31223/x5hj8m},
url = {https://eartharxiv.org/repository/view/15715/},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Energy and environmental impacts of rapid data-center expansion in the United Arab Emirates
AU - Ahmad Bin Thaneya
PY - 2026
DA - 2026/10/24
PB - EarthArXiv
DO - 10.31223/x5hj8m
UR - https://eartharxiv.org/repository/view/15715/
AB - Rapid growth in artificial intelligence (AI) is driving major data-center expansion globally, increasing demand on electric power and water-supply systems. Yet the energy and environmental implications of this growth remain comparatively understudied in hot, water-scarce regions such as the United Arab Emirates (UAE), where announced developments coincide with high cooling requirements, desalination-dependent water supplies, and a natural-gas-dominated electricity system. This study quantifies the electricity, water, and greenhouse gas (GHG) implications of UAE data-center expansion using a bottom-up capacity inventory, modeled operating loads, and a thermodynamic cooling model coupled to national electricity and water supplies. Scenario-aware Monte Carlo simulation propagates uncertainty across 2030–2040 operating scenarios with varying capacity expansion and clean-electricity generation levels. Median annual site electricity demand varies 6.62–37.8 TWh, reaching 11.5% of national generation. Attributed GHG emissions total 3.01–17.6 Mt CO2e annually under baseline grid conditions but fall 12–32% under alternative grids with increased clean generation. Expansion also reduces national clean-generation shares, requiring substantial additional solar photovoltaic capacity to meet UAE decarbonization objectives under some grid scenarios. When clean generation does not expand further to supply the added demand, increased natural-gas generation raises system-wide GHG emissions by 2.27–19.9 Mt CO2e (up to 22%) relative to a no-expansion reference. Water requirements remain high despite modeled power usage effectiveness near 1.22 for AI facilities. Median annual water demand is 10.6–63.1 Mm3, with desalinated cooling make-up comprising over 99% of demand. Dry-cooling configurations can reduce water demand by 98.2% but increase electricity demand by 11–13%, with the net electricity penalty roughly 20 times the desalination electricity avoided. These findings call for expediting clean-electricity deployment and planning water supplies alongside data-center expansion, given that low facility electricity overheads and lower attributed emissions from matching existing clean generation can coexist with substantial system-wide emissions and water requirements.
ER -