Skip to content
Review Open access

Energy and Water Conservation in AI Data Centers

Oct 2026 · The Journal of Young Innovators · 0 citations · 2 references

Abstract

As artificial intelligence applications become increasingly widespread, data centers' electricity and water consumption is drawing growing attention. This article analyzes 30 scientific and institutional reports to explore how these facilities can meet ever-increasing computing demands while conserving resources. The article compares published estimates with survey findings and uses a simple calculation to illustrate the impact of a lower power usage effectiveness (PUE) ratio. In 2023, electricity consumption at U.S. data centers reached 176 terawatt-hours (TWh); a 2024 national report projects that this figure will reach 325–580 TWh by 2028. Assuming IT power demand remains constant, reducing the PUE from 1.54 to 1.10 would decrease a facility’s total electricity consumption by 28.6%. Using a PUE of 1.40 as a baseline, total electricity consumption would decrease by 21.4% while achieving the same goal. The literature advocates improving both computational and facility operational efficiency; however, cooling choices must account for local water supply and the electricity source. Low-carbon electricity can reduce emissions, but it does not reduce electricity demand. The study concludes that resource conservation requires efficient data processing, appropriate cooling systems, and location-specific management of electricity and water resources. Beyond efficiency metrics, absolute consumption data are important for evaluating progress.

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.