Eco Survey flags strain on grids, groundwater due to AI data centres

India's AI data centre boom risks power grid volatility and groundwater depletion, prompting calls for urgent policy on energy and water use

The rapid expansion of artificial intelligence data centres can put severe stress on India’s grid systems and groundwater capacity, the Economic Survey has cautioned.

 

Asia’s third-largest economy now boasts a load capacity of 1.5–1.7 GW in 2025, which is expected to expand five-fold to 8 GW by 2030, according to research firm Jefferies. Almost 70 per cent of it is now being driven by hyperscalers.

 

“Data centres are also double-edged swords as they are very energy-intensive. With emerging hubs such as Malaysia (Johor), Japan, and Vietnam intensifying competition, addressing structural constraints such as energy shortages will be critical for India to position itself as a global AI data centre hub,” the survey said.

 

With AI development inextricably linked to physical infrastructure and data centres requiring large quantities of electricity, this can introduce volatility into power demand, posing risks to grid stability.

 

“Global experiences show that AI-driven data centre expansion can place significant strains on existing energy systems, even in advanced economies,” the survey added.

 

The huge demand, the JLL report said, will be driven by the growing adoption of AI technologies, as AI servers consume five to six times more power than traditional setups and require advanced liquid cooling systems. In addition, regulatory developments such as the Digital Personal Data Protection (DPDP) Act, 2023, and the RBI’s data localisation guidelines are driving enterprises to host and process data within India.

 

Despite generating nearly 20 per cent of the world’s data, India hosts only about 3 per cent of global data centres — around 150 out of 11,000 worldwide — according to Nasscom.

 

India’s data centre industry reached a major milestone in October when technology giant Google announced plans to set up a 1 gigawatt (GW) AI data centre in the port city of Visakhapatnam, Andhra Pradesh, investing $15 billion over the next five years.

 

AI data centres are specialised high-performance facilities designed to meet the intensive computational needs of machine learning and deep learning workloads.

 

But that has been criticised by advocacy group Human Rights Forum (HRF), which warns that the project will intensify groundwater depletion in a region where erratic rainfall and climate variability have already created acute water stress.

 

The most critical issue remains water — 25.5 million litres per year are needed for just a 1 MW load and computing infrastructure — and Indian cities may face operational disruption risks without government intervention. Experts cite the need for a data centre policy that accounts for water use in addition to other critical infrastructure support for land and electricity.

 

If India scales up AI data centres, it has the potential to add an extraordinary amount of stress on our already strained groundwater and freshwater reserves.