Why India's Electrostate Push Is Bigger Than You Think | Govindraj Ethiraj

ELI5 / TLDR

India just hit a record for electricity demand at the exact moment it was short on power. Pratik Agarwal, who runs three electricity businesses (a grid company, a renewables company, and a cable company), argues this is a turning point: India should become an “electrostate,” meaning more than half its energy comes from electricity instead of burning coal, petrol, and gas directly. The pieces are quietly falling into place — round-the-clock clean power is now cheaper than a new coal plant, India builds power lines faster than almost anywhere outside China, and the country can offer a data center a gigawatt of green power in two years, which the US currently can’t. The bottleneck isn’t generating power; it’s the wires that move it.

The Full Story

What “electrostate” actually means

An economy uses energy in two big buckets: electricity (lights, motors, ACs) and direct fuel-burning (a car burning petrol, a factory burning coal, a stove burning gas). Today only about 25% of India’s energy comes through the electricity socket. The rest is fire. An “electrostate” flips that — more than half flows through electricity. And of today’s 25%, only a quarter is renewable. So clean electricity is a quarter of a quarter of India’s energy. Lots of room to grow.

The trigger for the conversation: India’s peak power demand crossed 270 gigawatts in May, a record, and higher than government planners expected. The cause is almost comically simple.

“The country, of course, the killer app in electricity is ACs… It’s the peak summer, highest temperature and people have more ACs than they’ve ever had before.”

Agarwal’s framing is that a crisis is a wake-up call. COVID birthed UPI and digital payments; this energy crunch, he hopes, pushes India to electrify everything — cooking (induction instead of gas), transport (EVs instead of petrol), heavy industry (steel, cement, aluminum off coal).

The stakes, in his numbers: China grew its electricity use seven-fold over 2001–2021, roughly 9% a year. India is growing at 5–6%. If India could hit China’s 9% pace, electricity demand would grow 20x by 2047 — from 270 GW to 1,800 GW.

The clever bit: stop building power next to the factory

Here’s where it gets interesting. The old way to give a factory solar power was to build a solar farm on or near its premises. Agarwal’s company (Serentica Renewables) decided that doesn’t scale, for two reasons.

First, the spot next to your factory might have lousy sun or wind. Second — and this is the sharp insight — to deliver that local power you’d use the local distribution grid, which is owned by the same utility you’re trying to replace. Industrial customers are a utility’s most profitable accounts (it overcharges them to subsidize everyone else). Asking to use the utility’s own grid to poach its best customer was never going to go smoothly.

So they did something different: generate power wherever it’s cheapest — solar in Rajasthan, wind in Karnataka, pumped hydro elsewhere — and ship it across the national grid, which by law offers open access to anyone. Think of it like the difference between a restaurant growing vegetables in its own backyard versus buying from wherever the best produce grows and trucking it in on the highway.

“It’s like a symphony. You’ve got five, six instruments performing. Every 15 minutes something else is firing and you have to manage this entire symphony on a live continuous basis.”

That symphony is run by AI software (called Seran/Serna Nova) that predicts cloud cover, decides when to charge batteries, and matches generation to demand in real time. One concrete win from going national: combine wind from two states with different wind patterns, and you get more total hours of wind — which means you need fewer expensive batteries, which makes you cheaper than competitors.

The flagship example is Hindustan Zinc, the world’s largest zinc producer and a massive power user (800 MW). It was 100% coal-fired until 2021-22; by 2027 it’ll be 80% renewable — greener and cheaper at the same time.

The real bottleneck is transmission

Generating clean power is the easy part now. The hard part is the wires.

“If you talk to any renewable player anywhere in the world today and you ask them their number one constraint for growth, they will tell you the lack of grid capacity.”

Agarwal is genuinely proud of how India does transmission, and the reason is structural. Most countries run transmission as a single national monopoly paid on a “cost-plus” basis — you get paid your costs plus a margin, so there’s zero incentive to go faster or cheaper. India instead packages transmission projects and auctions them like highway contracts to the lowest bidder. Build it quicker and cheaper, earn a higher return. India now awards roughly $10–15 billion of transmission a year — the largest greenfield transmission market in the world.

But building a power line isn’t like building a road. For a road, the government buys the land. For a transmission line, you have to secure a continuous 50-meter corridor end-to-end — possibly across 5,000 separate landowners between Mumbai and Bhopal. So the real skill, oddly, isn’t engineering. It’s people.

The mismatch the interviewer presses on: renewables get built in a year, transmission takes 3–4 years. Agarwal admits it’s real and dangerous — if a renewable project can’t sell power in its first year, the lost revenue can wipe out its returns forever, and much of that money is foreign pension and sovereign capital that India badly needs to keep happy. The fix already underway: planners over-promised 24-month line delivery when 36 months was realistic, so renewable developers built ahead of the wires. New projects are now planned on honest 36-month timelines.

Drones, instant payments, and history graduates

Two ground-level details stand out. One: a labor shortage (infrastructure spending is 30-50% above all-time highs, and India uses 2x the labor of Thailand and 4-5x of Europe per kilometer of line). The answer is mechanization — drones now string the wires autonomously, talking to the tensioning machines.

The unexpected side effect: stringing wire the old way meant trampling a farmer’s crops. Drones fly over the top and don’t touch the crop. Farmers, delighted, become the project’s biggest ally instead of its biggest obstacle.

“If you’re not destroying their crop, they will be your best friend. They will help you with that project… It’s a very deep emotional connection.”

Two: this is why the company hires history graduates and runs a “stakeholder management academy” with a Wharton professor flying in twice a year. Building a power line is, in Agarwal’s words, not SpaceX — the engineering is routine, the contractor does the hard build. Success comes from knowing every village head (sarpanch) by name and having them on WhatsApp. A “fixing” mindset (bribery) is illegal and, he insists, doesn’t even work.

Why this is India’s moment for data centers

AI data centers are the new “killer app” for electricity. The US wants to keep that capacity at home but can’t — because it has no national grid, it can’t deliver huge blocks of power fast. India can.

“If somebody came to me as a data center company and said, can you do me a gigawatt of round-the-clock renewable power, I can say yes in 12 to 24 months in a location 2 hours from Mumbai… So India is in an extremely sweet spot.”

There’s a window — Agarwal pegs it as closing around 2028-29 — where India is one of the only places that can offer a gigawatt of cheap, fast, green power plus the land and grid connection as a plug-and-play package.

The economics that changed everything in October 2025

The quiet bombshell: in October 2025, Chinese battery prices hit a cost level that wasn’t expected until 2030-32. At that price, combining wind, solar, and batteries delivers round-the-clock renewable power cheaper than building a new thermal (coal) plant — even with coal prices falling. And since coal is an inflationary asset while renewables are flat-to-deflationary, renewables became the cheapest source of power anywhere in India, not just in pockets. Batteries plus solar can now do everything coal, gas, and nuclear do — 24/7.

What’s next for the grid

Three technical shifts: more HVDC (high-voltage direct current) lines to haul power long distances from remote Rajasthan solar with less loss; “FACTS” equipment so the grid can absorb solar power (the existing AC grid was built for the spinning mass of coal turbines, not for solar’s inverter-based electricity — India is “slightly behind” here); and far more battery storage placed on the grid rather than just at generation sites. His analogy: store grain in national highway warehouses so it can go anywhere, not just at the farm.

On money, Agarwal’s contrarian principle is no debt at the holding-company level — equity-funded, even at the cost of dilution. Singapore’s sovereign fund GIC and KKR are major backers. His logic: better to own 40% of a company that survives and grows to $10-20 billion than 100% of one that stays small and dies.

Key Takeaways

“Electrostate” = more than 50% of energy delivered as electricity. India is at ~25% today, of which only a quarter is renewable.
India’s record 270 GW peak demand is driven mainly by air conditioning during a brutal summer.
The scalable model for industrial clean power: generate where resources are best, ship over the national grid (open access), not build on-site using the incumbent utility’s local grid.
Combining wind from states with different wind patterns reduces the need for expensive batteries — the single biggest cost in round-the-clock renewables.
India auctions transmission like highway contracts (lowest bidder, build-operate-transfer), unlike the cost-plus monopolies of the US, UK, Europe, and Australia. This drives speed and lower cost.
India builds transmission concept-to-commissioning in 3.5-4 years vs 8-10 years in most of the world (China does ~3).
The core risk: renewables built faster than the wires to carry them. A renewable project that can’t sell in year one can lose its returns permanently.
Securing a transmission corridor means negotiating with thousands of landowners; farmers get a one-shot payment of roughly 50% of land value and keep their land.
Drone-strung wires don’t trample crops, turning farmers from obstacles into allies — and speed up construction.
October 2025: Chinese battery prices hit a 2030-level cost early, making 24/7 wind+solar+battery cheaper than a new coal plant.
India’s window to win global data center capacity (cheap, fast, green gigawatts) closes around 2028-29.
The existing AC grid struggles to absorb inverter-based solar; it needs FACTS equipment and grid-level battery storage to catch up.
India’s electricity market (~$100B/year) could double in six years at 6-7% growth.

Claude’s Take

This is a sponsor-adjacent interview — Agarwal runs the companies he’s praising (Sterlite/Serentica/Resonia), and the framing is relentlessly upbeat. Filter accordingly: “India does transmission better than the world” and “we make the highest IRRs in India” are the claims of a man selling a story to investors. The “electrostate” coinage is branding.

But strip the marketing and the substance holds up well, because the interviewer (Govindraj Ethiraj) actually pushes — on the transmission lag, on labor, on whether batteries can really replace baseload. The answers are specific and technically literate rather than evasive. The two genuinely non-obvious insights are worth keeping: (1) the reason India’s transmission is fast is the auction model versus cost-plus monopolies elsewhere — that’s a structural-incentives argument, not a patriotic one, and it’s persuasive. (2) The October 2025 battery cost inflection, if real, is the kind of quiet threshold-crossing that reorders an entire sector, and it squares with what’s broadly known about Chinese battery deflation.

The data-center “window closing in 2028-29” claim is the most speculative and the most self-serving, so hold it loosest. Score: 8 — dense, concrete, and you come away understanding why grids (not panels) are the real chokepoint of the energy transition, which is a genuinely useful mental model. Docked points only for the unavoidable conflict of interest.

Why India's Electrostate Push Is Bigger Than You Think | Govindraj Ethiraj | The Core Report