What Happened
- A new analysis projects that the rapid expansion of artificial intelligence workloads and data centres could increase India's peak electricity demand by up to 30 GW over the next few years — a significant addition to the existing peak demand of approximately 250 GW (FY25 record: 249.9 GW).
- India's data centre capacity is projected to grow from approximately 1.5 GW in 2025 to around 10 GW by 2030, driven by AI workloads, hyperscaler expansion (global cloud companies like Google, Microsoft, Amazon), and 5G deployment.
- Power demand from data centres is expected to reach approximately 57 terawatt-hours (TWh) by FY2030, raising the sector's share of national electricity consumption to 2.5–3%.
- Western and southern India — home to most global capability centres (GCCs) and hyperscaler campuses — are already seeing sharper power demand increases.
- The growth raises concerns about grid stability, water consumption (for cooling), and whether renewable energy capacity additions can keep pace with AI-driven electricity appetite.
Static Topic Bridges
India's Electricity Grid: Capacity, Peaks, and Grid Management
India's national grid is managed by the Power Grid Corporation of India (PGCIL) and operates through five regional grids unified since 2013 into one synchronous grid (excluding Sikkim and some island grids). Total installed capacity crossed 900 GW by early 2026 (including renewable energy targets under way). However, peak demand management remains a challenge — India's peak demand in FY25 reached a record 249.9 GW. Adding 30 GW of concentrated, 24/7 data centre load without corresponding flexible generation (dispatchable power) and grid upgrades would strain balancing mechanisms.
- India's peak demand crossed 240 GW for the first time in summer 2023; FY25 peak was 249.9 GW
- The grid operates through Regional Load Despatch Centres (RLDCs) and the National Load Despatch Centre (NLDC)
- Data centres are among the most power-intensive facilities — a hyperscale data centre can draw 100–500 MW continuously (equivalent to a small town)
- Unlike seasonal peaks (summer air conditioning), data centre load is constant (24×7), placing sustained pressure on baseload capacity
Connection to this news: A 30 GW addition from AI/data centres is not a peak demand challenge but a sustained baseload challenge — requiring India to add firm, reliable generation, not just intermittent renewables.
India's Data Centre Policy and Regulatory Landscape
India lacks a dedicated data centre regulatory framework but has enabled growth through infrastructure status grants and tax incentives. The Union Budget 2020 granted infrastructure status to data centres, enabling access to cheaper long-term funding. The National Data Governance Framework Policy (2022, revised 2023) and the Digital Personal Data Protection Act 2023 (DPDPA) shape data localisation requirements — key drivers for foreign hyperscalers to build data centres in India. India has earmarked ₹10,372 crore for semiconductors under the India Semiconductor Mission (ISM), indirectly boosting AI/data centre hardware supply.
- Infrastructure status (since Union Budget 2020) allows data centres access to priority sector lending and long-term bonds
- The Digital Personal Data Protection Act 2023 requires certain categories of data to be stored and processed in India — driving data centre construction
- Major hyperscaler investments announced: Google ($2 billion), Microsoft ($3 billion), Amazon AWS ($12.7 billion) — all in India, 2024–26
- India's data centre market is projected to reach $10 billion by 2030
Connection to this news: Policy incentives for data centres, combined with data localisation mandates, are accelerating capacity additions — making the 30 GW demand projection a near-certain trajectory rather than a speculative estimate.
Renewable Energy and the AI Power Dilemma
Globally and in India, AI companies have committed to powering data centres with 100% renewable energy. However, AI data centres require constant, uninterrupted power — a challenge for intermittent solar and wind. This creates the "AI power paradox": the more computing power AI requires, the more the grid is stressed, and the harder it is to meet clean energy commitments. Solutions being explored include co-location with nuclear power plants, large-scale battery storage, and green hydrogen for peaking.
- India's renewable energy target: 500 GW non-fossil capacity by 2030
- Solar and wind are intermittent — they cannot single-handedly power 24/7 data centre loads without storage
- IEA projects global data centre electricity consumption to double from 415 TWh (2024) to about 945 TWh by 2030 — largely AI-driven
- India's National Green Hydrogen Mission (2023) includes green hydrogen as a potential energy storage solution for data centres
Connection to this news: The 30 GW demand addition from AI/data centres collides with India's clean energy transition — requiring parallel investments in storage, transmission, and possibly nuclear power to meet both reliability and sustainability goals.
Key Facts & Data
- Projected AI/data centre addition to India's peak power demand: up to 30 GW
- India's FY25 peak electricity demand: 249.9 GW (record)
- India's data centre capacity: ~1.5 GW (2025) → ~10 GW (2030 projection)
- Data centre sector's share of national electricity: projected at 2.5–3% by FY2030 (~57 TWh)
- Global data centre power consumption: ~415 TWh (2024) → ~945 TWh (2030) — IEA estimate
- Hyperscaler investments in India: Google $2B, Microsoft $3B, Amazon $12.7B (2024–26)
- India's 500 GW renewable target by 2030 must now account for AI-driven baseload demand