Who should bear the cost of grid upgrades driven by data-centre growth? Analyse in the context of transmission-tariff design and consumer equity.
New · District CollectorCan you hold the district for 72 hours?Floods, a heatwave, an outbreak, a grievance queue. Take the decisions, see the consequences.Play →India's data-centre capacity grew about fourfold, from about 375 MW (2020) to about 1,500 MW (2025). Their power demand is projected at 13.56 GW by 2031–32 [1]. Grid investment is recovered through transmission charges paid by all users [2]. So the burden should be split by cause: share the costs of common shocks, and charge load added by one user to that user.
Breaking down the cost
- Global price shock: In the North-East, the estimated cost of the SCADA/EMS upgrade at State Load Despatch Centres rose from ₹521 crore to about ₹688 crore. AI-driven demand for hardware caused this, not local data centres [2].
- New connection load: Large, volatile AI loads need dedicated substations, lines and balancing. The network is planned to grow to 6.48 lakh ckm by 2032 [3].
- Benefit to the whole system: Modern control systems serve every user of the world's largest synchronous national grid [3].
Transmission-tariff design
- Pooled (postage-stamp) charges suit common upgrades like SCADA/EMS, because no single user caused the price shock.
- Causer-pays connection charges: dedicated assets for data-centre clusters can be recovered through capacity-based or "deep" connection charges.
- Flexibility incentives: lower charges or faster connection for centres that cut load when the grid signals stress. This reduces the investment needed for peak demand.
- Green obligations: The government plans to meet the extra demand mainly through renewables [4]. That needs renewable-plus-storage contracts covering every hour of use, not just a match over the whole year.
Consumer-equity concerns
- Unfair spreading of costs: households in regions that host no data centres end up paying for a global AI boom [2].
- Regional imbalance: small north-eastern utilities feel a ~32% cost jump more than large states would [2].
- Counterpoint: data centres used only about 1.5% of global electricity in 2024 [5]. Charging them too much could discourage digital investment.
Cost allocation should therefore follow causation. The global hardware shock should be shared, with central support such as the Power System Development Fund for small states. Dedicated network costs should be charged to data centres, with discounts for flexible load. A tariff of this kind, framed by CERC under the Electricity Act, 2003, serves SDG 7 and SDG 9 while keeping to the equity goal of Article 38. India's digital growth should not be paid for by its poorest consumers.
Sources
- 1Data centre capacity in the country has increased from about 375 MW in 2020 to around 1500 MW by 2025 — Government of India press release (Digital India)capacity growth; 13.56 GW demand by 2031–32
- 2AI data-centre boom puts power grid under pressure, The Hindu (news report, 11 Oct 2026)SCADA/EMS cost ₹521 crore → ₹688 crore; recovery through transmission charges; burden on North-East utilities
- 3India's power transmission network crosses 5 lakh circuit km — All India Radio News (Govt. of India)world's largest synchronous national grid; transmission expansion to 2032
- 4Data Centres in India: Infrastructure for the Digital Age (PIB, Sept 2026)extra data-centre demand to be met mainly through renewables
- 5IEA, *Energy and AI* (hosted on NITI Aayog ICEMF).pdf) — data centres about 1.5% of global electricity use in 2024