What will power AI data centres?

Why in News ?

• India’s electricity demand, stagnant at around 5% annual growth for two decades, is now rising rapidly due to new high-energy sectors.
• Key demand drivers: AI & Data Centres, Electric Vehicles (EVs), 5G/IoT, green hydrogen, and digital economy expansion.
• The US, China, and Big Tech firms are already witnessing 25%+ annual power demand surges from AI data centres.
• India is planning GW-scale AI data centres (Google at Visakhapatnam, Reliance at Jamnagar) and exploring Small Modular Reactors (SMRs) as clean, reliable energy sources for them.
• The Union Budget 2025 launched a ₹20,000 crore Nuclear Energy Mission to add 100 GW nuclear capacity by 2047, including SMRs.

Relevance:

• GS 3 – Energy, Infrastructure, and Technology: AI-driven electricity demand, sustainable energy mix, nuclear innovation through Small Modular Reactors (SMRs), and linkage with the IndiaAI Mission.
• GS 2 – Governance and Policy: Inter-ministerial coordination between MeitY, MoP, and DAE for energy–technology convergence; clean energy policies under Budget 2025; regulatory reforms for private participation in nuclear energy.
• GS 3 – Environment: Low-carbon power strategy, Net Zero 2070 alignment, and sustainable infrastructure for digital economy expansion.

Background: India’s Power Demand Trends

• Past 20 years: Electricity demand grew at ~5% annually — relatively stable due to efficient grids, low industrial expansion, and moderated population growth.
• Shift (post-2023): Rise in data traffic, EV charging, AI computation, and green hydrogen manufacturing expected to double electricity demand by 2030.
• India’s per capita electricity consumption (2025): ~1,350 kWh — 1/3rd of global average (~4,000 kWh), but projected to rise steeply.
• Planning challenge: Aligning digital economy growth with sustainable, low-carbon electricity expansion.

Why India Needs Data Centres

• Digital India Mission, data localisation laws, and explosive data usage demand domestic storage and processing capacity.
• Current capacity: 1.4 GW
  • vs. Europe: 10 GW
  • India has 2× more internet users than Europe, yet 1/7th capacity.
• Expected growth:
  • By 2027: 2–3× increase (to ~4 GW).
  • By 2030: >5× increase (to ~7–8 GW) with AI and LLM infrastructure.
• Drivers:
  • Data privacy & localisation mandates.
  • 5G and IoT ecosystem.
  • Cloud computing, fintech, and generative AI expansion.

Power Demand from AI Data Centres

• Traditional server racks: 15–20 kW.
• AI/LLM GPU racks: 80–150 kW (≈6× higher load).
• Global data centre electricity usage:
  • 2024: ~460 TWh
  • 2030 (projection): ~1,000 TWh
  • 2035: ~1,300 TWh (~6% of global generation).
• Case studies:
  • China:
    • Data centre electricity use to reach 400+ billion kWh by 2025 (~4% of total power).
    • CAGR ~18% (2023–2030).
  • US (Dominion, Virginia):
    • Electricity and peak demand projected to rise >25% in 5 years due to data centres.

Data Centre Hubs: Global and Indian

Global

• US: 51% of global capacity — hubs in Texas, Virginia, Ohio, Phoenix, Wisconsin, Pennsylvania.
• Other nations: China, Norway, UK, Germany, Japan, Malaysia investing in AI-grade infrastructure.

India

• Emerging AI data centre clusters:
  • Visakhapatnam (Google) – GW-scale, AI-optimised.
  • Jamnagar (Reliance Industries) – part of IndiaAI Mission.
  • Mumbai, Chennai, Bengaluru, Hyderabad – existing hyperscale hubs (Yotta, AdaniConneX, Sify, CtrlS).
• IndiaAI Mission (2024):
  • Focus on indigenous AI models, large-scale compute infrastructure, and clean energy linkages.

Powering the AI Era: Energy Mix Options

Renewables

• Solar, wind, and hydro as clean options but intermittent and storage-dependent.
• Storage (battery, pumped hydro) still developing — costly for 24/7 AI operations.

Natural Gas & Green Hydrogen

• Used as backup for renewables ensuring grid reliability.
• Hydrogen blending and onsite generation emerging in industrial clusters.

Emerging Alternatives

• Geothermal energy (pilot projects in Ladakh & Gujarat).
• Nuclear fusion research under ITER collaboration (long-term).

Small Modular Reactors (SMRs) – The Key Innovation

• SMRs emerging as reliable, low-carbon baseload solution for AI data centres.
• Range: 1–300+ MW capacity.
• Advantages:
  • Modular, factory-built → faster deployment.
  • Passive safety systems → no human/manual intervention needed.
  • Can be located near consumption hubs → no transmission losses.
  • Flexible for remote/industrial sites.

Global Investment & Regulation in SMRs

Investment Landscape

• Total global SMR investment: $15.4 billion
  • $10 billion – public funding
  • $5.4 billion – private capital (tech & energy companies)
• Big Tech (Google, Microsoft, Amazon) exploring SMR power purchase deals for AI facilities.

Regulatory Reforms (Global Trends)

Six key areas of SMR regulation evolving internationally:

1. Technology-neutral frameworks (beyond large LWR models).
2. Streamlined licensing – combined construction-operating licences.
3. Fleet-wide approvals – enabling standardised mass deployment.
4. Factory certification – for modular manufacturing.
5. Risk-informed requirements – proportional safety zones.
6. International harmonisation – via IAEA standards & mutual recognition.

Leading Regulatory Models

• U.S. ADVANCE Act (2024) – accelerates SMR licensing.
• Canada – Vendor Design Review (pre-licensing pathway).
• UK – Regulatory sandbox approach.
• IAEA – Nuclear Harmonization and Standardization Initiative (NHSI).

India’s SMR Push

Budget 2025 Initiatives

• ₹20,000 crore outlay under Nuclear Energy Mission.
• Target: 100 GW nuclear capacity by 2047.
• IndiaAI–Nuclear synergy: Aligns AI infrastructure growth with clean baseload energy.

Key Developments

• BARC’s BSMR-200 – 200 MW Pressurised Heavy Water Reactor (PHWR) variant.
• 55 MW SMR for remote areas in isolated grid mode.
• Holtec–India partnership for technology transfer.
• Private participation reforms:
  • Planned amendments to Atomic Energy Act (1962) & Civil Liability Act (2010).
  • Aims to attract $26 billion private investment.

State-level Role

• Pre-approval of coal plant sites for SMR conversion.
• Land facilitation, safety training, and workforce reskilling.
• Demonstration projects integrated with green hydrogen hubs.

SMR Safety and Environmental Aspects

• Passive safety features:
  • Natural convection cooling.
  • Automated shutdown systems.
  • Accident-tolerant fuels withstand higher temperatures.
• Waste & transport regulation:
  • Need for new frameworks addressing factory fabrication, transport risks, and spent fuel disposal.
  • HALEU fuel (high-assay low-enriched uranium) requires specific waste management protocols.
• IAEA support:
  • SMR Regulators’ Forum.
  • Safeguards by Design Programme – balancing safety, economics, and security.

Opportunities for India

1. Energy Security – 24/7 baseload for AI infrastructure.
2. Climate Goals – Low-carbon transition aligned with India’s Net Zero 2070 target.
3. Export Potential – India can become SMR exporter to Global South via cost-effective indigenous tech.
4. Industrial Repurposing – Utilize decommissioned coal plant sites.
5. Employment & Skill Creation – Reskill coal workforce for nuclear operations.

Challenges Ahead

• Regulatory delays – outdated laws not suited for SMRs.
• Public perception & safety concerns.
• Financing barriers – high upfront capital cost despite modularity.
• Waste disposal & liability – still unresolved.
• Grid integration – ensuring SMR–renewable hybrid stability.

Core Takeaway

India’s next energy transition will be driven not just by renewables, but by AI-driven demand.
Data centres and SMRs together define the digital–nuclear nexus of the future — where clean, constant power meets the data economy, enabling India’s journey from Digital India to Energy-Secure India by 2047.

---