AI Data Centres in India: Energy Demand & Challenges

Rapid expansion of AI-driven data centres is reshaping India’s electricity demand profile.
AI infrastructure requires continuous, high-quality, zero-downtime power, unlike traditional industrial or household consumption.
As India pushes data localisation and Digital Public Infrastructure (DPI), energy governance faces new structural stress.

Why AI Data Centres Are Energy-Intensive

Always-on Operations: AI model training, deployment, and real-time updates require a 24×7 electricity supply.
Rising Global Demand: The IEA estimates that data centre electricity consumption has grown by 12% annually since 2017.
- It is expected to reach ~945 TWh globally by 2030.
Load Quality: Data centres need stable voltage and frequency, not just aggregate capacity.

Data Localisation Mandate: The Digital Personal Data Protection (DPDP) Act, 2023, along with sectoral mandates from the RBI and SEBI, has made domestic storage of sensitive data mandatory.
- Data sovereignty has emerged as a market-shaping force, compelling global cloud providers to invest locally.
Digital Demand Explosion: Growth in AI workloads, cloud computing, fintech, e-governance, OTT platforms, gaming, and IoT has created sustained demand for high-capacity, low-latency infrastructure.

United States: Tech firms exploring nuclear power to ensure reliable baseload for AI infrastructure.
China: Expansion of nuclear capacity and development of Small Modular Reactors (SMRs) to support energy-intensive digital clusters.
These trends contextualise India’s SHANTI Bill, 2025, aimed at advancing nuclear energy.
- It seeks to incentivise private sector participation in the nuclear power sector to reach 100 GW capacity by 2047 and Net Zero by 2070.

Fragmented Governance: Power planning is split across generation, transmission, and distribution.
- Electricity regulated by multiple institutions (Centre, States, CERC, SERCs).
Sectoral Silos: Data centres governed by MeitY while Land, water, and environmental approvals handled separately.
- Furthermore, poor coordination between digital planning and grid capacity assessment.
Outdated Planning Logic: Focus on aggregate capacity addition rather than reliability and load quality.
- Also, limited incentives for nuclear, hydro, or storage-backed renewables.

Grid Stress: Continuous high-load demand strains local distribution networks.
- India’s installed data capacity is projected to reach 2.5 GW by 2027.
Coal Dependence: Despite 235.7 GW of non-fossil capacity, coal remains the primary baseload stabiliser. This increases emissions and complicates climate commitments.
Water Intensity: A 100 MW data centre consumes ~2 million litres of water per day.
- Major hubs (Mumbai, Chennai, Hyderabad) already face water stress. AI-related water demand globally may reach six times Denmark’s annual usage.

No National Data Centre Policy: Absence of a comprehensive national data centre, while the Draft Data Centre Policy (2020) has not been formally adopted.
- State-level policies prioritise investment incentives over sustainability.
Weak Environmental Regulation: No mandatory energy-efficiency or water-use norms and sustainability standards largely voluntary.
Institutional Capacity Deficit: Regulators lack expertise to assess long-term digital loads.
- Energy governance and data governance are treated as separate domains.

With the DPDP Act setting the foundation, policy must now enable execution by streamlining approvals through single-window clearance mechanisms.
Grid modernisation should be prioritised to support continuous, high-quality power for digital infrastructure.
Skill development must be treated as a national strategic investment; a reliable workforce is as critical to data centre resilience as physical infrastructure.

The focus has shifted from land acquisition to strategic securing of power and water resources.
New facilities must be AI-ready by design, incorporating liquid cooling, energy efficiency, and scalability from inception.
Long-term renewable energy procurement and retention of highly skilled operations teams will be decisive competitive differentiators.

Cloud and data centre partners are now integral to business continuity, legal compliance, and ESG commitments.
Corporate sustainability and compliance risks are increasingly interlinked with infrastructure partners’ practices.