NextFin News - India's solar boom has run into a physical wall, and the problem is not a shortage of panels. The country added roughly 24 GW of solar capacity between October 2025 and April 2026, lifting the national fleet to about 154 GW. Yet at midday in March 2026, solar and wind briefly supplied 41% of all electricity generated, up six percentage points in a single year. The grid was never built to absorb a midday peak like that. Coal plants, which still provide almost all of India's balancing flexibility, are being forced below the floor at which they can operate safely. The system's response is to throw clean power away.
The question for investors and policymakers is no longer whether India can build more solar. It is whether batteries can be deployed fast enough, and cheaply enough, to keep that solar from going to waste. The early evidence points both ways: tenders are surging, but the cheapest bids are so aggressive that analysts warn a large share may never be built.
The Mechanics of a Grid Running Out of Room
India's renewable problem is a timing problem, not a volume problem. On an annual average basis, solar and wind supply around 17% of electricity generation. At midday, that share can reach 41%. The rest of the system, dominated by coal, must therefore swing from near-full output at night to its lowest point around noon, every single day.
Coal has a hard operating floor. Most plants cannot run stably below about 55% of their rated capacity, a threshold known as the minimum technical load. Once a coal unit reaches that floor, it cannot reduce output further, so it can no longer provide downward reserves — the capacity to absorb excess generation when the sun is brightest. When coal hits its floor, renewable output must be curtailed simply to keep the thermal fleet physically operable.
That constraint is now binding with measurable frequency. In April 2026, the system required coal to generate below 55% of rated capacity in more than half of midday dispatch intervals. In March, the coal fleet was already running an average of 3.2 GW below its technical minimum during midday hours. By April, renewable curtailment accounted for 37% of all down-regulation, or 816 million units, up from near zero a year earlier. Across fiscal 2025-26, keeping coal above its technical minimum required the curtailment of 2.1 TWh of renewable generation, equivalent to roughly ₹629 crore of foregone electricity. At its worst, up to 6% of solar and wind generation was cut for this reason alone — not because of transmission congestion or weak demand, but because coal could not go lower.
The real-time market signal tracking this stress confirms the trend. TRAS-Down volumes, the grid's instruction to generators to reduce output, roughly quadrupled from around 600 GWh to more than 2,100 GWh between September and November 2025. Growth accelerated again from March 2026, with more than 1,400 GWh added in just two months. On individual days in early May, volumes exceeded 120 GWh. Peak-hour curtailment returned to 4% of solar and wind generation by April 2026, with a headroom deficit of 3.2 GW — comparable to levels seen the previous November, even though April sits outside the most seasonally constrained window.
Here lies the structural heart of the problem: the same curtailment numbers are now being produced by a considerably larger solar fleet. The constraint has not changed. The amount of solar pressing against coal's operating floor has.
Why the Fix Is Batteries, and How Much Is Enough
The arithmetic of the fix is surprisingly modest at the margin. Analysis by Ember estimates that around 10 GWh of storage, charging during the midday window, would have been sufficient to absorb the surplus renewable generation, hold coal above its minimum technical load, and avoid curtailment. The role of that storage is primarily to provide downward reserves — to be dispatched so coal plants can run closer to their technical minimum rather than being displaced by solar.
But "enough for today" is a moving target. The Central Electricity Authority, in its National Electricity Plan, projects India will need 411.4 GWh of total energy storage by 2031-32 — split between 236.2 GWh of battery energy storage systems and 175.2 GWh of pumped-hydro storage — to support 364 GW of solar and 121 GW of wind. For the nearer term, the requirement is 82.37 GWh by 2026-27. The gap between the 10 GWh that would clear today's surplus and the 236 GWh of BESS the planner says is needed by 2032 measures how fast the problem compounds as solar keeps growing.
The evening side of the equation is equally unforgiving. Solar generation tapers off just as demand holds firm, and the National Load Dispatch Centre has flagged potential shortfalls of 15 to 20 GW during those evening hours. In April 2026, India recorded its highest-ever electricity demand of 256 GW as temperatures surged; solar contributed nearly 58 GW during peak afternoon hours before falling away toward evening. A solar panel cannot move its own output across that daily canyon. A battery can.
The second-order effect is where the market has not fully priced the risk. The first-order consequence of curtailment is wasted generation today. The second-order consequence travels through project finance: a solar developer's power purchase agreement is priced on an assumed number of delivered kilowatt-hours, and every unit the grid throws away widens the gap between contracted revenue and actual cash flow. If curtailment becomes structural rather than occasional, lenders will reprice that risk across the entire renewable book, lifting the cost of capital for new solar even as panel prices keep falling. Value therefore migrates away from pure generation and toward flexibility — the batteries, the dispatch rights, and the grid connections that determine which megawatt-hours actually get paid for. The market has priced a record solar buildout. It has not priced the flexibility premium that the buildout itself makes unavoidable.
Deployment itself is not the binding constraint — speed is achievable. The 3.37 GWh Khavda BESS project was commissioned in May 2026, within ten months. As of March 2026, around 1.8 GWh of grid-scale BESS was installed, with the majority coming online in the final six months of the fiscal year. Cumulative tendered energy storage capacity has surged from 6.8 GW in 2018 to 90.7 GW in 2025. Standalone ESS tenders — storage contracted without being tied to a specific renewable asset — accounted for more than 71% of total capacity tendered in 2025, with standalone BESS making up 60% of that. In February 2026 alone, roughly 3,070 MWh of standalone BESS capacity was auctioned across state and central procurers.
The Bidding War That Threatens to Undo the Boom
Enthusiasm in the auction room has outrun project economics. Tariffs in standalone BESS tenders fell sharply through 2025, with the lowest discovered tariff reaching INR 1.48 lakh per MW per month for two-hour systems, against an indicative benchmark of INR 2.3 lakh per MW per month. Nearly 75% of allocated two-hour capacity now sits in what analysts classify as the "risky" category — a significant gap between discovered tariffs and actual project costs. Viable outcomes have largely been confined to early-stage, smaller-scale procurements in Karnataka, Tamil Nadu, Telangana, and Gujarat.
"The surge in standalone storage tenders has coincided with declining battery prices and supportive policy measures such as the introduction and expansion of viability gap funding for standalone BESS projects."
Vasu Mor, research associate at JMK Research and Analytics and co-author of a May 2026 report on tariff viability, said as much. But the same report, produced with the Institute for Energy Economics and Financial Analysis, warns that implementation delays of up to 18 months may persist because of financial closure, procurement, and commissioning challenges, and that cost pressure at lower tariffs could compromise asset quality.
The cancellation data underscores the execution risk. Around 5,300 MWh of standalone BESS capacity and nearly 2,800 MW of firm dispatchable renewable energy and storage capacity currently stand cancelled. Aggressive bidding is a familiar pattern in India's renewable auctions: BESS prices fell 22.3% between the TERI 20 MW Delhi auction in September 2023 and the GUVNL Phase 3 250 MW tender. When the lowest bid wins regardless of deliverability, the clearing price becomes a forecast of future distress.
"Although the near-term challenges may lead to some project cancellations or delays, the eventual growth of ESS is inevitable. This momentum is already visible, with the majority of the around 1.8 gigawatt-hour of grid-scale BESS capacity installed as of March 2026 having come online in the last six months of FY2026."
Prabhakar Sharma, senior consultant at JMK Research, said. The technology mix is also being tested. Heavy reliance on lithium-ion chemistry has exposed the sector to global supply-chain shocks, and tendering agencies are expected to shift toward alternative technologies with longer lifespans and lower raw-material risk.
"Going ahead, the BESS technology landscape will be a diversified mix of storage technologies including Li-ion, flow batteries, sodium-ion etc."
Charith Konda, energy specialist for India mobility and new energy at IEEFA, said. Commercial signals are already appearing. Reliance New Energy energized a 5 MW/50 MWh vanadium flow battery in Gujarat in September 2024, the country's first utility-scale non-lithium long-duration storage system. Tata Power commissioned a 100 MW/200 MWh standalone BESS in Rajasthan under a 25-year PPA at ₹5.85 per kWh, and Adani Energy Solutions brought a 40 MW/120 MWh system online in Gujarat at ₹5.95 per kWh.
The Counter-Case: Why Batteries Might Not Be the Only Answer
The strongest argument against a battery-centric reading is that India has other levers. Coal plants can be retrofitted for more flexible operation, allowing them to cycle deeper without hitting their floor. Pumped-hydro projects offer long-duration storage without lithium exposure. Transmission expansion can move surplus power across regions rather than curtailing it locally. On this view, batteries are one option among several, and the market's underbidding may simply reflect competition among substitutable sources of flexibility.
There is also a demand-side argument. Data centres, electric-vehicle charging, and industrial electrification are adding new load, some of which could be shaped to absorb midday surplus. If demand grows faster than solar, the surplus could shrink on its own.
Neither argument holds up as a substitute for storage at the scale now required. Coal flexibility has a hard physical limit: once a unit is at its minimum technical load, no retrofit creates more downward room, and the midday coal share has already been pushed to just over 50% from nearly 90% at night. Pumped hydro is geographically constrained and slow to permit; the CEA's own plan assigns it 175 GWh of the 411 GWh total, leaving 236 GWh for batteries. Transmission helps only if a neighboring region has spare capacity in the same hour, but the midday solar peak is national and synchronized, so surplus arrives everywhere at once. And demand growth, while real, does not solve the timing mismatch: the evening shortfall of 15 to 20 GW occurs precisely when solar output collapses and thermal capacity is least flexible.
The falsifying signal is specific and observable. If midday curtailment falls over the next two fiscal years while solar additions continue and little BESS is commissioned, the coal-flexibility thesis would be vindicated and the battery-urgency case weakened. The opposite is more likely. With the solar fleet still growing and the October–November 2026 window approaching, curtailment in those peak hours is expected to exceed 2025 levels unless storage comes online at scale first.
What Comes Next: Beneficiaries, the Exposed, and the Watchlist
The beneficiaries of the storage imperative are clear. Developers with execution capacity and balance-sheet depth — NTPC, JSW Energy, Tata Power, Adani Energy Solutions, ACME, and Hero Future Energies — are best positioned to convert tenders into operating assets. JSW Energy has committed to a 29.6 GWh locked-in pipeline and a 40 GWh target by 2030, framing storage as "a foundational infrastructure investment, not an add-on." Equipment suppliers and domestic manufacturers stand to gain if localization policies hold, though India remains heavily dependent on imported cell technology. The market itself is small but expanding fast: India's BESS market was valued at about $2.05 billion in 2026 and is projected to reach $8.59 billion by 2031, a 33.2% compound annual growth rate.
The exposed parties are the bidders who won on price alone. With nearly three-quarters of allocated two-hour capacity in the risky tariff band and thousands of megawatt-hours already cancelled, the next 18 months will separate bankable projects from paper awards. Procurers — state distribution companies and central agencies such as SECI and NTPC — face the risk of paying for capacity that never materializes, or of renegotiating at higher tariffs after delays. Ratepayers ultimately bear the cost either way: through higher tariffs for delivered storage, or through continued curtailment of power they have already contracted to buy.
Three signals are worth watching. First, the commissioning cadence: whether the roughly 6 GWh of BESS expected online by the end of 2026 actually connects on schedule. Second, tariff discipline: whether future auctions introduce cost-reflective floors after the 2025 underbidding episode. Third, the midday curtailment share through the October–November 2026 peak window — the first real stress test of whether storage is arriving ahead of the problem or chasing it.
The forward view splits into three scenarios. In the base case, storage deployment tracks the CEA's trajectory with delays concentrated among the riskiest 2025 bidders; curtailment rises through the October–November 2026 window before new BESS capacity begins to absorb it in 2027. The upside case requires two things to break right: auctions introduce cost-reflective tariff floors quickly, and the roughly 6 GWh expected by end-2026 connects on schedule — under that path, the 10 GWh needed to clear today's surplus arrives before the next winter peak and curtailment stabilizes. The downside case is a repeat of the solar-auction cycle of the 2010s, when aggressive bids produced a wave of renegotiations and stalled projects: if more than half of the risky two-hour capacity is cancelled or delayed beyond 18 months, curtailment could climb well above the 4% peak-hour level already recorded, and the cost of capital for the entire renewable complex would rise with it.
Split by horizon, the picture differs sharply. In the short term, sentiment and auction headlines will dominate, and execution stumbles will be read as sector-wide weakness. Over the medium term, fundamentals will sort winners from losers as projects either deliver power or default. Over the long term, the structural math is unforgiving: India's 500 GW non-fossil target by 2030 cannot be met with solar alone, because a grid that cannot time-shift energy cannot use all the energy it generates.
India's solar boom was built on falling panel prices and ambitious targets. Its next phase will be built on something harder: the ability to move a megawatt-hour from noon to night. Without batteries at scale, the country risks building the world's cheapest solar fleet and then paying to turn it off.
Explore more exclusive insights at nextfin.ai.
