On May 1 this year, the electricity price on India’s real-time power market fell to zero for two 15-minute blocks. It was a cool holiday with record solar output, but such lows are spreading. Exchange prices fell below five paise a unit 31 times between April and June, against six times a year earlier. After sunset, they can reach Rs 10 a unit, the highest price the regulator allows.
Cheap at lunch, maximum price at dinner. That daily swing says more about India’s energy future than any capacity milestone.
The milestones are real. India took more than a decade to go from 1 gigawatt (GW) of solar power to 100 GW, then reached 168 GW by August 31 this year. Yet grid operators keep telling solar and wind plants to cut output. This is called curtailment. India is throwing away sunlight it has already paid to catch.
Why electricity cannot wait
Start with simple physics. A power line is a pipe, not a tank. Rice can sit in a godown for months, and water can wait in an overhead tank. Electricity in a wire cannot wait. Every unit produced must be used somewhere at that same moment.
Imagine a cyclist on a flat road. If the road tilts downhill, the bike speeds up. The grid behaves similarly. When supply exceeds demand, its frequency rises above the normal 50 cycles a second. When demand exceeds supply, frequency falls. Move too far either way, and equipment begins to disconnect to protect itself.
So when the wires are full, the operator may have to tell a solar farm to stop sending power. The panels still sit in full sun, but their energy never reaches anyone.
The number that misled us
Last month, headlines said India had curtailed 118.6 GW of renewable capacity in September. That sounds as if much of India’s clean power vanished. It did not. A gigawatt is like a car’s speed, while a unit of electricity is like the distance covered. The 118.6 GW is the capacity affected. The energy actually lost was 378.52 million units, or 1.28 per cent of renewable generation.
That may sound small, but by a yardstick from the think tank Ember, it equals one day’s electricity for more than five crore urban middle-class households. September was also a milder month, 24 per cent below a year earlier. Curtailment peaked at 1,166.52 million units in July, and JMK Research says it rose fourfold between May and August compared with last year. The Centre for Research on Energy and Clean Air (CREA) found that about 4.2 per cent of solar output was cut in July and August.
There is also a hidden cost. Mint, citing Ember, reports about 40 GW of mostly solar capacity without signed sale agreements, partly because of weak transmission and a lack of storage. New renewable awards fell from 47.3 GW in 2023-24 to 14.7 GW in 2025-26. A small monthly percentage may hide a slowdown in the next wave of projects.
Three mismatches
The trouble comes down to three mismatches: of place, of speed and of time.
The first is place. Gujarat and Rajasthan produced nearly 79 per cent of September’s curtailment, and their solar power must travel to reach many of its users. Between January and March, Ember counted about 300 million units lost to transmission limits: 178 million in the north, 122 million in the west and none in the south. The sun was not the problem. The route out was.
The second is speed. A solar or wind project can often be ready in 12 to 18 months. A high-voltage line must get past land disputes and forest clearances. India has met only about 80 per cent of its yearly transmission targets over the past five years, and one in four major inter-state schemes is a year or more late. About 20 GW of projects expected this financial year could wait over four months for a connection.
The third, and most important, is time. Solar power peaks around midday, but India also needs large amounts of electricity after dark. On May 21, the day-ahead price sat at the Rs 10 cap from 6:30 pm, a paper from the Prime Minister’s economic advisers shows. CREA reported power shortages alongside the curtailment of July and August, while coal stocks at plants fell to about 36 per cent of requirement by September 30.
The missing tank
If a wire is a pipe, a battery is the missing tank. India’s battery capacity is growing fast and reached about 9.3 gigawatt-hours by June. At this year’s record demand of 270.8 GW, that could run the country for about two minutes. Batteries need not do that alone, but the sum shows the scale. The Central Electricity Authority projects a need for 236 gigawatt-hours by 2032.
The Authority’s draft rules would require new ground-mounted solar and onshore wind projects commissioned after July 1, 2027 to carry two hours of storage, sized at 10 per cent or more of their capacity, rising to four hours from 2029. It is a good start and a modest one. A 100 MW solar farm would first need 20 megawatt-hours, while on a sunny day it could produce over 20 times that.
Ember estimates that stored renewable power can be delivered at Rs 7-8 a unit, below the Rs 9-10 many states pay for peak power. It also estimates that 3-4 GW of two-hour batteries could absorb most of the power lost on temporary grid connections.
Why every reader should care
Exchange prices do not set your electricity bill directly. Only about 13 per cent of India’s power trades there. But distribution companies still buy costly power for peak hours, and such costs land in tariffs, subsidies or the weak finances of state utilities. The air pays too, whenever coal fills an hour that stored sunlight could have served.
India’s target of 500 GW of non-fossil capacity by 2030 counts what is installed, not what is delivered. A country can hit that number and still burn more coal than it needs to.
It would be wrong to call this a crisis. The grid has not collapsed, and batteries are arriving. That is exactly why this is the moment to act, while the problem is still small enough to fix at a reasonable cost.
Three steps would help. First, publish delivered clean power alongside installed capacity. Second, let batteries work as part of the transmission system at congested points, as Ember suggests. Third, make noon power cheap for users who can shift, such as water pumps, cold stores, factories and electric vehicles.
The sun has kept its side of the bargain. A panel counted in an official report is only a promise. It becomes electricity only when its power reaches a home, at the hour that home needs it. India has learned how to catch sunlight. It must now learn how to deliver it.
The author is a physicist at the University of North Carolina at Chapel Hill and contributor opinion writer at the Wall Street Journal; Views presented are personal.