AI data-center energy and water use projections to 2030

Data center energy & water projections for 2030 (AI-driven) and mitigation strategies

Electricity demand

Global (IEA, Energy and AI, 2025): Global data‑centre electricity consumption is projected to roughly double to ~945 TWh by 2030 (Base Case), reaching just under 3% of total global electricity. Demand grows ~15% per year from 2024–2030 — more than four times faster than the rest of the electricity sector. AI is the main driver: electricity use by accelerated (AI) servers grows ~30% per year and accounts for nearly half of the net increase. The US and China account for ~80% of the growth (US +~240 TWh, +130%; China +~175 TWh, +170%). Beyond 2030, the Base Case rises to ~1,200 TWh by 2035.

United States (LBNL / DOE, 2024 U.S. Data Center Energy Usage Report): US data‑center electricity use climbed from 58 TWh (2014) to 176 TWh in 2023 (4.4% of US electricity) and is projected to reach 325–580 TWh by 2028, or ~6.7%–12% of total US electricity (growth of 13%–27% per year), driven largely by AI servers.

Water consumption

• IEA (Energy and AI): Total data‑center water consumption was ~560 billion liters in 2023 (≈two‑thirds indirect/power‑generation‑related, ≈one‑quarter direct cooling), projected to rise to roughly 1.2 trillion liters by 2030. A typical 100‑MW US data center uses ~2 million liters/day.
• UNU‑INWEH (UN University Institute for Water, Environment and Health, 2026): Under a scenario where AI reaches 40% of global data‑center electricity by 2030, AI‑related water consumption could reach ~9.3 trillion liters per year — enough to meet the basic domestic water needs of >1.3 billion people in Sub‑Saharan Africa for a year. The same report estimates AI‑hardware e‑waste could reach ~2.5 million metric tons/year by 2030.

(Note: figures vary widely by scope — direct vs. total, AI‑only vs. all data centers — explaining the gap between the IEA and UNU numbers.)

Proposed mitigation strategies

• Efficiency gains (hardware, software, infrastructure): IEA's High Efficiency Case projects >15% energy savings versus the Base Case from stronger efficiency progress.
• Cooling technology shifts: Moving from evaporative ("swamp") cooling — which evaporates ~80% of drawn water — to closed‑loop / liquid cooling cuts direct water use by ~70%–90%; air cooling conserves water but raises electricity use (cooling is 20%–40% of a data center's energy), a key energy‑water trade‑off.
• Transparency / disclosure mandates: UNU‑INWEH calls for governments to require AI providers to disclose energy, water, emissions and resource‑efficiency footprints.
• Demand‑side / user behavior: preferring low‑footprint tasks (text over image/video), concise prompts, batching and reuse of results, and provider prompts when a request is resource‑intensive.
• Circular‑economy measures: recycling, component reuse and e‑waste reduction for AI hardware.

Sources:

• https://www.iea.org/reports/energy-and-ai/energy-demand-from-ai
• https://www.energy.gov/articles/doe-releases-new-report-evaluating-increase-electricity-demand-data-centers
• https://eta-publications.lbl.gov/sites/default/files/2024-12/lbnl-2024-united-states-data-center-energy-usage-report_1.pdf
• https://earth.org/9-3-trillion-liters-of-water-un-report-exposes-unfathomable-footprint-of-data-centers-as-ai-booms/
• https://computeforecast.com/blogs/9-3-trillion-liters-of-water-un-report-exposes-the-growing-environmental-cost-of-ai-infrastructure/
• https://www.eesi.org/articles/view/data-centers-and-water-consumption