India’s Data Centers Drive Severe Water Stress in Jamnagar & Thane

Jamnagar and Thane in India have been pinpointed as global hotspots for water stress, directly linked to the burgeoning expansion of data centers driven by Artificial Intelligence (AI) infrastructure. Rystad Energy forecasts a significant increase in cooling water demand within these already water-scarce regions.

Key Water Consumption Projections

• Direct water consumption by data centers stood at 222 billion litres annually in 2025.
• This is projected to nearly triple to 644 billion litres annually by 2030 if current water-saving measures fail to keep pace with AI-led growth.
• An aggressive water-efficiency strategy could limit annual consumption to approximately 388 billion litres by 2030.
• A moderate scenario would see consumption reach around 543 billion litres.
• Regions experiencing high or extremely high water stress are projected to account for 34% of the data center sector’s global direct water consumption by 2030.
• Mandating the use of the least water-intensive cooling technologies in these areas could reduce consumption by as much as 45%.

Globally, the direct water consumption by data centers is set to escalate dramatically. This surge is predicated on an AI-led growth in computing capacity that current water-saving measures may not adequately address. Implementing strategic cooling technology and design choices offers considerable potential for water savings.

The Mechanics of AI Cooling Demand

Minh Khoi Le, global head of data center and hydrogen research at Rystad Energy, highlighted the critical role of cooling technology as AI workloads expand. AI servers generate substantially more heat than traditional computing equipment. This increased heat necessitates advanced solutions.

• Advanced cooling solutions include rack-level liquid cooling, which reduces strain on facility-level systems.
• This also enables greater adoption of dry cooling.
• Dry cooling can save approximately 2.15 litres of water for every kWh of IT load.
• However, dry cooling requires an additional 0.30-0.74 kWh of electricity, potentially shifting the water footprint to the power generation system.
• In the US, indirect water consumption from electricity supply can more than double direct consumption.

Operator Efficiency Varies Widely

Water efficiency varies significantly among different operators and locations, showcasing the impact of climate and cooling architecture. These differences reflect diverse operational strategies and environmental conditions.

• Reported average direct-site water-use effectiveness (litres/kWh) in 2025:
• AWS: 0.12 (average), ranging from 0.02 in Stockholm to 2.85 in Jakarta.
• Meta: 0.19
• Microsoft: 0.27
• Digital Realty: 0.59
• Equinix: 0.91

The projections underscore a critical challenge for urban centers like Jamnagar and Thane, particularly as AI infrastructure continues its rapid expansion. Urgent adoption of the least water-intensive cooling technologies in these vulnerable regions could significantly mitigate the impending water crisis.