Data Centers in Indiana and the Midwest: Assessment of Power and Water Demand in the AI Era
インディアナ州と中西部におけるデータセンター：AI時代の電力と水需要の評価 (AI 翻訳)
Sherif Attallah, Jennifer Warrner
🤖 gxceed AI 要約
日本語
本研究は、インディアナ州を中心に中西部の大規模データセンター開発の電力と水需要を評価。AIワークロードの増加により、単一キャンパスで数十TWhの追加需要が発生する可能性があり、冷却技術の選択が直接水使用に大きな影響を与える。間接的な水消費も重要であり、統合的なエネルギー・水ネクサス計画と持続可能性報告の必要性を強調する。
English
This study assesses the power and water demand of hyperscale data centers in the US Midwest, focusing on Indiana. It finds that a single campus can add tens of TWh of electricity demand, with water impacts heavily dependent on cooling technology. Indirect water use from electricity generation remains significant, underscoring the need for integrated energy-water nexus accounting and transparent corporate reporting.
Unofficial AI-generated summary based on the public title and abstract. Not an official translation.
📝 gxceed 編集解説 — Why this matters
日本のGX文脈において
日本でもAI需要増加に伴いデータセンター建設が急増しており、電力調達や冷却技術の選択がGX戦略の重要な要素となっている。本論文の水・エネルギー統合評価の枠組みは、日本の事業者や自治体が持続可能性報告や設備計画を策定する上で参考になる。
In the global GX context
Globally, data center electricity demand is projected to nearly double by 2030, driven by AI. This paper highlights the critical need for integrated resource planning and transparent disclosure of both direct and indirect water use, which aligns with TCFD and ISSB frameworks emphasizing climate-related risks and opportunities.
👥 読者別の含意
🔬研究者:Provides a quantitative framework for assessing energy-water nexus in data centers, useful for sustainability and infrastructure researchers.
🏢実務担当者:Data center operators and corporate sustainability teams can use the findings to evaluate cooling technologies and clean energy procurement strategies.
🏛政策担当者:Highlights the need for utility integrated resource plans and municipal infrastructure planning to accommodate exponential load growth from AI.
📄 Abstract（原文）
The Midwest is experiencing a rapid surge in hyperscale data-center development, largely driven by the growing computational demands of artificial intelligence (AI) workloads. Centering on Indiana while incorporating regional comparisons, this study compiles a verified public dataset of active and announced projects, models their annual electricity consumption across load-factor scenarios, and estimates direct on-site water requirements under alternative Water Usage Effectiveness (WUE) assumptions. Results reveal that a single multi-gigawatt campus can contribute tens of terawatt-hours (TWh) of additional annual electricity demand, while water impacts depend heavily on the choice of cooling technology, ranging from evaporative to dry or hybrid systems. Even when direct on-site water use is minimized, the indirect water footprint embedded in electricity generation remains significant, highlighting the urgent need for integrated energy-water nexus accounting within utility Integrated Resource Plans (IRPs), municipal infrastructure planning, and corporate sustainability reporting. Recent studies converge on projections of exponential load growth from AI operations: global data-center electricity use, estimated at approximately 415 TWh in 2024, is expected to nearly double to 900-1,000 TWh by 2030, with AI emerging as the dominant driver (Takci, 2025). Although many hyperscale operators are adopting dry or hybrid cooling systems that nearly eliminate direct water withdrawals, the indirect consumption associated with thermally based electricity generation continues to pose major sustainability challenges, particularly within grids characterized by high fossil or nuclear generation shares. These findings reinforce the need to couple WUE disclosure with transparent reporting of grid composition and to accelerate the transition toward 24×7 clean-energy procurement strategies.
🔗 Provenance — このレコードを発見したソース
- openalex https://doi.org/10.29007/m2wlfirst seen 2026-06-21 05:08:45
🔔 こうした論文の新着を逃したくない方は キーワードアラート に登録（無料・3キーワードまで）。
gxceed は公開メタデータに基づく研究支援データセットです。要約・翻訳・解説は AI 支援で生成されています。 最終的な解釈・検証は利用者が原典資料に基づいて行うことを前提とします。