India's Power Grid Is Being Pushed Closer to Its Breaking Point

Peak electricity demand hit a record 270.82 GW in May 2026. The grid held by day, but evening shortfalls and state-level outages show how thin the margins are.

**270.82 gigawatts.**

That is the most electricity India's grid has ever had to supply at a single moment. The Ministry of Power recorded it at **3:45 pm on 21 May 2026**, during a punishing heatwave. It was the fourth record in four days: demand stood at 257.37 GW on 18 May, 260.45 GW on 19 May and 265.44 GW on 20 May, before breaking through 270 GW. In July, demand crossed 270 GW twice more.

A gigawatt (GW) is a thousand megawatts, about the combined output of one or two large coal-fired generating units.

The record did not come out of nowhere. India's peak demand was about **184 GW** in 2019-20 and about **203 GW** in 2021-22. It crossed **250 GW** in May 2024. The 2026 figure is roughly **47%** higher than the peak six years ago. The grid met this year's peak, but the week it was set also brought power cuts in several states. Both facts matter.

Several forces are pushing in the same direction.

**Heat.** Summer peaks are driven above all by cooling: air conditioners, coolers and fans. The Ministry of Power linked the May records directly to cooling load during a heatwave, with temperatures above 45°C across much of north and central India. The government has also cited El Niño, a Pacific weather pattern that tends to bring hotter, drier conditions to India, as a reason for higher demand this year. Between April and June 2026, peak demand ran about **12%** higher than a year earlier.

**Hotter nights.** It is not only afternoons. When nights stay hot, air conditioners run late into the evening and after dark, when solar power is gone. That is where the system is most stretched, as later sections show.

**Growth and electrification.** More homes own appliances, more villages have longer supply, and more of the economy runs on electricity, from factories and metros to electric vehicles. The government says average rural supply rose from **12.5 hours** a day in 2013-14 to **22.6 hours** in 2025-26. Every extra hour of supply is extra demand the grid must meet.

**A weak base year.** Part of the jump is also a rebound. In 2025, a mild, rainy summer kept the peak to about **243 GW** in June, well below the government's own forecast. Total electricity requirement in 2025-26 grew less than 1%. The 2026 heat brought back demand that had been deferred.

At the moment of the record, coal-based and other thermal plants supplied **62.8%** of electricity. Solar supplied **22%**, hydro **5.8%** and wind **5%**. The mid-afternoon timing helped: solar was near its daily high.

That contribution is growing fast. In July 2026, India recorded its highest-ever output from solar and wind together, **103.7 GW**, of which solar was about 75 GW. On 29 July 2025, renewable sources met more than half of the country's demand for the first time.

But the afternoon peak is no longer the hardest moment for the grid. The evening is.

Late on the evening of 21 May, hours after the record was met, national grid operator Grid-India reported a peak **shortfall of about 2.57 GW**. That is demand that could not be met at that moment, and it translated into power cuts in some areas.

The reason is simple. Solar power fades after sunset, but cooling demand stays high. The grid then relies on coal, hydro and gas plants to fill the gap. The government told Parliament that the most coal and lignite plants have supplied at once is about **188.8 GW**, roughly **75%** of total generation of 251.4 GW during non-solar peak hours.

Power exchanges show the same pattern in prices. On at least 10 days in May, the spot price of electricity on India's largest power exchange fell below **10 paise a unit** in some daytime slots, because solar power was abundant. At night, prices on those same markets hit the regulatory ceiling of **₹10 a unit**. In late April, in the separate high-price market, where costlier power from gas and imported coal is sold, prices touched the **₹20 ceiling** during non-solar hours.

In short, India increasingly has plenty of power at noon and not quite enough at 10 pm. Storage would help, but most of it is not built yet.

On 22 May, the power ministry appealed to consumers to use electricity "wisely and judiciously". In several states, people were already living with outages.

**Uttar Pradesh** was the most visible case. The state's power corporation attributed its problems largely to the shutdown of **11 major thermal power stations** during May. Between 15 and 22 May, between **2,638 MW and 4,529 MW** of generating capacity was unavailable each day. On 20 May, rural areas faced nearly **seven hours** of cuts. The state's consumer council chief said many rural districts were getting only 5 to 8 hours of supply, against a norm of 18. Even ruling-party legislators raised the issue publicly.

Residents also reported night-time outages of 40 minutes to an hour in parts of **Chennai**, complaints in **Delhi** and **Noida**, and protests over longer cuts in parts of **Odisha**. Officials in several places blamed local faults, overloaded transformers and heat-related breakdowns rather than a lack of power.

That distinction is important. National figures can show supply meeting demand while local networks fail. For a household sitting in the dark, it makes little difference whether the cause was a national shortfall or a burnt-out transformer.

On paper, India has far more capacity than it needs. Installed generating capacity reached **548.86 GW** by the end of June 2026. Non-fossil sources, mainly solar, wind, hydro and nuclear, made up **54.18%** of it.

But installed capacity is not the same as power available at the peak. Solar plants produce nothing at night. Wind varies with the weather. Some coal plants are down for maintenance or breakdowns at any time. What matters for the evening peak is "firm" capacity, meaning power that can be switched on when needed.

The government points to record additions. In 2025-26, India added **64,695 MW** of generating capacity, the most ever in a single year, including **50,905 MW** of renewable energy. India also met its target of getting **50%** of installed power capacity from non-fossil sources in 2025, five years ahead of its 2030 commitment. In March 2026, it updated its climate pledge to aim for **60%** by 2035.

It is also still building coal. The Central Electricity Authority (CEA), the government's technical planning body, projects a need for about **315 GW** of coal and lignite capacity by 2035-36, up from roughly 229 GW in March 2026. About **21,080 MW** of thermal capacity has been commissioned since April 2023, and **47,545 MW** is under construction. The government argues that coal remains the backbone for evening and night demand until storage catches up.

Storage is the missing piece. Batteries and pumped-storage hydro plants can soak up cheap solar power at midday and release it in the evening.

The plans are large. The CEA estimates India will need about **174 GW** of storage, holding **888 gigawatt-hours** of energy, by 2035-36. As of June 2026, about **15,754 MW** of battery storage was under construction, a further 11,747 MW had been awarded, and 19,192 MW was out to tender. About **15,870 MW** of pumped storage was under construction.

Actual commissioned capacity under government support is still small. The first viability gap funding scheme for battery storage, which offers public money to make projects commercially viable, now covers **13.8 GWh** after battery prices fell. By July 2026, only **0.5 GWh** of it had been commissioned and **₹168 crore** paid out. A second scheme for **30 GWh**, with **₹5,400 crore** in support, was launched in June 2025. The CEA has also advised that new solar projects include storage equal to at least 10% of their capacity, for two hours.

The gap between plans and working batteries is where the evening risk sits. Each summer that passes before storage arrives at scale is a summer in which the evening peak rests on coal plants running close to their limit.

Two other risks sit behind the headline.

**Coal stocks.** As of 18 July 2026, thermal plants held **41.4 million tonnes** of coal, enough for about 13 days at high output. The government says supplies are adequate. Gas makes up less than **2%** of generation, which has limited the direct effect of the West Asia conflict on power supply.

**Transmission.** Power has to move from where it is made, often solar-rich Rajasthan and Gujarat or coal-rich central and eastern India, to where it is used. Inter-regional transmission capacity has reached **1,20,340 MW**. India's grid now runs as a single synchronised national system. But local distribution networks, the wires and transformers that reach homes, are where many summer failures actually occur.

Electricity is a shared responsibility of the Centre and the states, and it is the state distribution companies, or discoms, that buy power and deliver it to consumers. Their finances decide whether they can buy expensive power in a crunch or keep local networks in good repair.

There has been genuine improvement. Aggregate technical and commercial losses, which cover power lost to theft, faulty meters, weak wires and unpaid bills, fell from **22.62%** in 2013-14 to **15.04%** in 2024-25. The gap between what it costs discoms to supply a unit and what they recover narrowed from ₹0.78 to about **₹0.06** a unit. Discoms as a group reported a small profit of about **₹2,701 crore** in 2024-25 after years of losses. Old unpaid bills owed to power generators fell from about **₹1.39 lakh crore** in June 2022 to about **₹3,300 crore** by March 2026, under late-payment rules introduced in 2022.

But the balance sheets remain heavy. Discoms had outstanding borrowings of about **₹7.26 lakh crore** as of March 2025. Tamil Nadu's discoms owed over **₹1 lakh crore**, Rajasthan's about **₹98,500 crore** and Maharashtra's about **₹90,700 crore**. In six states, about **₹2.74 lakh crore** of debt has been disallowed by regulators from recovery through tariffs, which means it cannot simply be passed on to consumers.

A national record says little about whether any particular state is managing its supply well. Uttar Pradesh supplied a peak of more than 31,000 MW in May even as its rural consumers faced long cuts. The useful questions are state-level: did the discom contract enough power for the peak, were its plants available, and did its local network hold?

**What does "peak demand" mean?** It is the highest amount of electricity drawn from the grid at any single moment, measured in gigawatts. It is different from total consumption over a day or year, which is measured in units (kilowatt-hours).

**If there is enough power nationally, why do cuts happen?** Cuts can come from a local fault, an overloaded transformer, a state not buying enough power, or plants in that state breaking down. National figures average these out.

**Why is the evening harder than the afternoon?** Solar power, now a large part of the supply, fades after sunset while cooling demand stays high. The gap must be filled by coal, hydro, gas or storage.

**Will electricity bills go up?** Tariffs are set by state regulators, not by the national peak. But when discoms buy costly power on exchanges during shortages, those costs can later feed into tariffs or state subsidies.

**Evening shortfalls.** Grid-India publishes daily data on peak demand met and shortages. Watch the evening shortfall figure during heatwaves, not just the headline daytime peak.

**Battery commissioning.** Watch how much of the 15,754 MW of battery storage under construction is actually switched on before summer 2027.

**Thermal plant availability.** Watch forced outages of coal plants in states such as Uttar Pradesh during the next summer peak, and whether maintenance is scheduled around the hot months.

**The next demand forecast.** The CEA's projections and states' 10-year resource adequacy plans will show whether planned capacity keeps pace with demand of around 270 GW and beyond.

**Discom finances.** Watch the next annual rating of distribution utilities and whether current dues to power generators start rising again.

**Exchange prices at night.** Frequent evening prices at the ₹10 ceiling would show that supply after sunset is still tight.

A record peak is not in itself good or bad news. It reflects a growing economy with more people able to afford cooling, and a warming climate that makes cooling essential. India's grid met the record in the afternoon, largely thanks to solar power and coal.

The strain shows up elsewhere: in evening shortfalls, in state networks that buckle during heatwaves, and in a storage build-out that is still mostly on paper. Whether the grid is near breaking point depends less on the national record than on what happens after sunset and at the level of each state.

**Note on perspective**: {{Based on Ministry of Power statements and replies in Parliament, Central Electricity Authority plans, grid operator and power exchange data, and reported state-level outages as of 25 September 2026. The piece treats the record as a test of system readiness and focuses on evening supply and state discoms, where accountability lies.}}