What India Can Learn from the US About Data Centres
Last week, the Department of Energy began consulting with data center developers, transmission utilities and communications companies as it worked on a roadmap to meet the industry’s rapidly growing electricity demand and initiate dialogue on reliable supply, dual grid connectivity, green energy and grid stability. The government’s own predictions are striking: internal planning at the Department of Energy shows that electricity demand from data centers will reach 26 GW in 2032 and 35.7 GW in 2040; This represents an almost ninefold increase in ten years. The Ministry has just started drawing the map. This article explains what the map should include.
India’s total operational data center capacity exceeds approximately 1.5 to 1.7 GW by late 2025; cumulative long-term investment commitments are projected to exceed $100 to $125 billion by 2030-32 (CBRE India Real Estate Reports; JLL Data Center Intelligence). However, the sector still lacks a binding national policy framework. India still lacks a single, comprehensive national framework to manage the energy, water and grid impacts of data center expansion. Growth, like green hydrogen before it, has been shaped state by state through incentives and bilateral agreements, through Maharashtra, Tamil Nadu, Uttar Pradesh and Telangana. The only thing a data center can’t get along with is physics. The pressures underlying this rise deserve to be named clearly.
The reason for this lies in the way data centers consume electricity. Three prints stand out.
1. Brutal base load. A 100 MW data center operating continuously will consume approximately 876 GWh per year in IT load alone. Cooling and other facility systems contribute significantly to the electricity required beyond this IT load. Unlike most industrial loads, this demand operates at full intensity around the clock with zero tolerance for interruption.
2. Geographic clustering. Mumbai alone holds 45% to 50% of India’s total data center capacity through undersea cable landings. The top four hubs – Mumbai, Chennai, Bengaluru and NCR/Hyderabad – account for over 85 to 90% of the total tier 1 colocation supply (JLL India Data Center Report; CBRE India). These are cities already strained by the demands of urbanisation, and developers in Navi Mumbai have reported interconnection lead times of 18 to 36 months for private 100MW+ substation installations (MSEDCL; Data Center Association of India) and the queue is growing.
3. Clear the energy bottleneck. The problem is not producing green megawatts. India is doing this. The real challenge is to provide solid power to a particular substation at 3am on a windless night. Intermittent renewables alone cannot solve this.
America preview
The United States is not, from afar, a cautionary tale. This is a vivid demonstration of where India is heading, and the numbers make uncomfortable reading.
As of April 2026, Texas grid ERCOT was tracking approximately 410 GW of large load interconnection demand, approximately 87% of which came from data centers (ERCOT Senate Committee Presentation, April 2026). In the first quarter of 2026 alone, 198 GW of major load was applied to the interconnect (Ascend Analytics, May 2026), more than double the grid’s all-time peak demand and occurring in a single quarter. PJM capacity prices rose sharply as projected demand growth, including for data centers, outpaced new supply and reached the regulatory limit. Last week, New York became the first U.S. state to impose a moratorium: Governor Kathy Hochul signed Executive Order 62 on July 14, 2026, pausing state environmental permits for up to a year for new hyperscale data centers of 50 MW or more and pursuing legislation that would repeal sales tax exemptions (Governor Hochul’s Office, July 14, 2026). By May 2026, approximately 12 GW of data center load demand was already in the NYISO interconnect queue; More than 8 GW had entered in 2025 alone, as noted in the Executive Order.
Public sentiment has changed sharply. Only one in three Americans now approve of the pace of data center construction and just 14% would be happy to see one built in their community, according to a Reuters/Ipsos poll (June 2026), and dozens of state legislatures are considering similar measures. India has a chance to avoid this altogether by planning better in advance. This advantage should not be wasted.
Four lessons India cannot ignore
A. Ask for flexibility from day one
Each new connection should be conditioned on interruption capacity from the outset, paired with time-of-day tariffs that carry real incentives. A data center that reduces 10 to 15% of its load in the evening hours does not create a burden on the grid. It is an entity. This needs to be built in from the initial approval, not added later when the network is already struggling.
B. Embrace the off-grid data center, but design it better than America
This is where India’s greatest opportunity lies. S&P Global Commodity Insights and Wood Mackenzie estimate over 50 GW of planned behind-the-meter gas generation tied to hyperscale data centers across ERCOT, PJM and MISO (S&P Global Market Intelligence). Approximately 2 GW is currently operational; these include xAI’s Colossus in Memphis; this facility became operational in mid-2024 using up to 35 mobile gas turbines with a capacity of approximately 420 MW, bypassing the local service queue and becoming operational approximately 100 days after the decision (Global Energy Monitor; Greater Memphis Chamber) and the Stargate campus in Abilene, Texas.
The business logic is simple. AI data centers running Nvidia H100 or B200 chips can earn $10 to $12 million per MW per year in gross infrastructure hosting revenues (Bernstein Research; Morgan Stanley AI Infrastructure Reports). The fact that it takes two years to be operational instead of eight easily outweighs the cost of producing electricity on site. Speed is what the job requires.
The important nuance here is that the American build is optimized for speed, not efficiency. Since the turbines will run out years in advance, developers give away everything they can get their hands on. This is where India’s opening lies: Take the model and build it better.
Modern lean-burn gas engines, such as the Wärtsilä 34SG or Jenbacher J624, provide 42 to 48% electrical efficiency, far beyond simple cycle turbines fielded throughout the United States (Wärtsilä Technical Data Sheet; Jenbacher Specifications). They maintain this efficiency at partial load, synchronizing in about a minute and reaching full output in a few minutes. Gas turbines lose up to 15 to 20% of their capacity in Indian summer temperatures, while reciprocating engines lose less than 5% power in the same conditions (ASME Digital Collection). By pairing engines with heat recovery to drive coolers, trigeneration takes 30 to 40% of the cooling load from the primary power source (US DOE Combined Heat and Power Technical Resources) while increasing total fuel use beyond what any turbine installation can achieve. In a country where cooling is required year-round rather than seasonally, this is not a marginal gain. This is a structural situation.
India also has two advantages that the United States does not have. First, India has approximately 24.8 GW of installed gas-based generation capacity; of these, 14 to 18 GW are idle or operating well below capacity, lacking affordable fuel, but sites, infrastructure and expertise are already available (Central Electricity Authority; IEEFA, July 2025). The easing in LNG prices is currently leading to the reactivation of around 17 GW of this idle capacity as data center demand accelerates (Reuters, March 2026). Secondly, Article 9 of the Electricity Act 2003 provides a clear legal framework for dependent generation, allowing a data center developer to establish on-site generation without a generation licence, subject to applicable requirements under the Act and the Electricity Rules 2005 (Department of Energy). The regulatory pathway that US developers have to create from scratch already exists in Indian law.
C. Deliberate centralization
India must put its policy weight behind data center and edge computing investments currently taking shape in Bhubaneswar, Lucknow and Guwahati, where state governments in Odisha, Uttar Pradesh and Assam have signed MoUs with providers such as Nxtra, STT GDC and RailTel (STT GDC India; MeitY Data Center Policy Framework). Data Center Economic Zones, in tier 2 locations selected for network headroom and water availability rather than just fiber access, will distribute the load more evenly across the national network. Northern Virginia illustrates the risks of allowing data center growth to be heavily concentrated in a single region. A district with such a large share of national capacity is the cost of leaving geography entirely to market forces.
D. Reasonable planning and conditional access
The U.S. is now moving toward dedicated bulk load tariffs, minimum commitment obligations, and cost allocation rules that ensure data centers pay their fair share of the infrastructure they need. New York went further with an outright moratorium. India does not need to get to this point, but it does need to make energy efficiency standards, water norms and energy self-sufficiency connectivity conditions right from the start and not as fixes applied after the industry is established.
The regulatory power India wields today is exactly what the United States gave up in its boom years and is now trying to take back. The right time to use this is before the moratorium starts to look like the only option.
The window for action is now.
India has a real advantage here; the ability to watch another major market take the fast route, face the consequences, and still have time to make different choices. The data center boom is already here. The Ministry of Energy’s grid roadmap is a real and necessary step. However, a road map is only as useful as the rules written into it and the solid power capacity behind it. The wave is here. The question is whether the framework will be ready when it arrives.
This article is written by Raman Chopra, CEO and all-time Director of Caparo Power Limited.
