Supply, stress and shortfalls: How India met its high electricity demand in September 2026

by Upasa Borah and Renuka Sane.

India has been touching new electricity demand peaks this year: 256.1 GW on 25th April, 265 GW on 20th May, and an all-time peak of 270.8 GW on 21st May. This demand has also been higher than the levels projected by the Central Electricity Authority (CEA). As demand grows, the ability of the system to provide sufficient supply when it is needed becomes increasingly important.

September 2026 offers a useful window into how the system balances demand and supply under pressure. Despite higher electricity generation relative to the previous year and high utilisation of the coal fleet, electricity shortages were the highest recorded since August 2023. The power exchange was also stressed during the month: average market clearing prices rose by 105% compared with September 2025, while purchase bids increased by 281% and sell bids fell by 45%. Over 100 captive power plants have been ordered to run at full capacity from October and sell surplus power on the exchanges to meet rising demand. Reports of critically low coal stocks at several coal power plants have added to concerns about supply for the coming months.

Against this backdrop, we examine electricity demand, supply and shortages at the transmission level in September 2026, defining shortages as the gap between electricity requirement and available supply. The article is descriptive: it sets out how the system behaved during the month rather than offering causal analysis. Establishing how demand, supply and shortages moved through the month gives a factual baseline against which causes and policy responses can be assessed. We ask three questions:

1. When did shortages emerge, and how did they vary across the days over the month?

   Daily and intra-day variations show how the system responded to high demand, and identify points at which it came under stress. This high-frequency analysis provides a snapshot of the state of the system, offering a useful starting point for understanding the operational challenges that are to become more important as electricity demand and renewable generation grow.

2. How did the power system respond to changes in demand?

   1. What was the composition of supply and its variation on the days of highest and lowest demand?
   2. How intensively were different generation sources utilised?
   3. How much renewable generation was curtailed?

   Installed capacity alone does not tell us how much electricity is available to meet demand at any given time. This depends on how much different sources can generate, how quickly they can respond to changes in demand and whether electricity can be delivered where it is needed. In the absence of storage, solar generation can meet a substantial share of daytime demand but falls as the sun sets. Thermal generation then needs to increase output, often quickly, though it can create technical and cost constraints. Hydro, oil and gas generation and storage can help manage these fluctuations, so their availability and responsiveness also matter. Solar and wind generation can get curtailed at times even when the plants are capable of generating more electricity due to transmission constraints limiting evacuation, or because the system has more power than it can absorb. Understanding these patterns is important for assessing how the power system manages both periods of scarcity and surplus.

3. How was the power exchange used during periods of high and low demand?

   Short-term markets also show how the system responds to changes in demand. Most of India's electricity is procured through long-term power purchase agreements, but buyers can buy additional power on the country's three exchanges. These offer segments across different time horizons, including the Day-Ahead Market (DAM), where electricity is traded a day before delivery, and the Real-Time Market (RTM), where trades take place an hour before. Such near-real-time markets help buyers respond to unexpected shifts in demand and generation, but their usefulness depends on how much electricity is offered and at what price. Both the DAM and RTM have a regulated price ceiling of Rs 10/kWh, which is often hit during high demand, leaving markets to clear at volumes well below what buyers bid for. A well-functioning market matters because it lets prices signal scarcity and guide decisions by generators and buyers.

Data sources

The study uses the following data sources:

1. Grid-India's Power Supply Position Reports: These reports are compiled by Grid India using data reported by State Load Dispatch Centres to Regional Load Dispatch Centres, which are then compiled at the national level. They provide information on energy supplied, energy shortages, peak demand met during solar and non-solar hours, and shortages during those periods. These values are measured at transmission buses and do not capture behind-the-meter generation like rooftop solar, or distribution-level constraints and outages. Our analysis thus focuses on system-level supply adequacy rather than delivery to end consumers or more localised grid constraints.

   The power supply position reports define shortage as the difference between unrestricted requirement and energy available, as per the Central Electricity Authority (Furnishing of Statistics, Returns and Information) Regulations, 2007. Energy availability implies supply at ex-bus, while unrestricted requirement is the sum of available energy, frequency correction, scheduled and unscheduled load shedding. The precise definitions and methodologies of these terms, however, could not be located, and it has been flagged in some Regional Power Committee meetings that there are inconsistencies in the methodologies adopted by states and a lack of transparency in the estimation of shortages. The results presented in this article should thus be interpreted with this caveat and a closer examination of the methodology underlying these estimates remains an important avenue for future work.

2. National Power Portal Daily Generation Reports: These are used to analyse how the generating fleet was operating and include data on daily capacity utilisation of thermal, nuclear and hydro power plants. Capacity utilisation here is defined as the ratio of a day's maximum generation to the available capacity on line, where available capacity is the installed capacity minus the capacity under outage. It is worth noting that these are daily aggregates rather than hourly measures. An hourly or 15-minute interval measure would provide a better indication of how much generation capacity was actually available to meet demand at different points in the day, but such data was not available.

3. Grid India's Daily VRE Report: These reports are used to measure the quantum of solar and wind generation that was curtailed daily. Data for 28, 29 and 30 September were not available.

4. India Energy Exchange's Day Ahead Market (DAM), Real-Time Market (RTM) and High Price Day Ahead Market (HP-DAM): These provide daily 15-minute blockwise data on aggregated purchase and sell bids, market-clearing quantities and market-clearing prices.

Demand and shortages in September 2026

India's electricity requirement in September 2026 stood at 1,62,736 GWh. Of this, about 545 GWh went unmet, implying a shortage of around 0.3% of total demand. These shortages varied across days and hours, but occurred on almost every day of the month, with 25 September being the only exception. Figure 1 shows the daily electricity required and met. Daily demand averaged around 5,424 GWh, with the highest requirement of 5,791 GWh recorded on 9 September; 51 GWh of this went unmet. The highest daily shortage was 54 GWh, recorded on 11 September. As demand declined towards the end of the month, shortages also fell.

Figure 1: Daily energy required and met

Peak demand refers to the highest quantum of electricity drawn from the grid at a point in time. There are two peaks generally observed, one during solar hours of 6 am to 6 pm, and another in the remaining non-solar hours. Figure 2 shows the daily peak demands during solar and non-solar hours, along with the shortage during these peaks. With the exception of a few days towards the end of the month, non-solar peak demand was lower than the solar-hour peak demand. Shortages, however, occurred during the non-solar peaks. On 9 September, for instance, the peak shortage was 7.72 GW at around 7:06 pm.

Figure 2: Solar and non-solar peak demand and shortages

Shortages emerging in non-solar hours, while the system met its highest solar-hour demand, is a defining feature of a grid with a growing share of solar. India's solar generation has increased significantly over the years, allowing the system to meet high levels of daytime demand. But this also leaves the question of how to meet demand as solar generation declines in the evening, particularly when demand remains high or increases.

Composition of electricity supply

Thermal plants generate around 71% of India's electricity, and non-fossil sources (renewables, nuclear and large hydro) the remaining 29%. The mix shifts by time of day and season. In September, coal accounted for an average of 65% of total generation, followed by renewables at 17% and hydro at around 11%.

When it comes to grid balancing, supply matters more in real time rather than in daily or monthly aggregates. We, therefore, study two days in September: 9 September, the day of highest electricity requirement, and 27 September, the day of lowest requirement. Figure 3 plots demand and supply at 15-minute intervals on the two days: On 9 September, solar accounted for 23% of total generation during solar hours (6 am to 6 pm), reaching a peak share of 33% at 1:15 pm when it generated 84.25 GW. On 27 September, solar accounted for 21% of generation during solar hours, reaching a peak of 30% at 1:45 pm when it generated 61.8 GW.

Figure 3 also plots the net load (dark solid line), defined as demand minus generation from solar and wind. The resulting profile shows the duck curve: as solar generation increases during the day, net load declines, reducing the amount of generation required from other sources. As solar generation wanes in the evening, while demand often stays high or increases, net load rises, and other sources must ramp up supply to meet the gap. This pattern is visible on both days, but on 9 September, higher demand meant that the evening ramp-up began earlier, around 2 pm onwards. Gas and hydro generation also increased during the morning and evening periods of lower solar generation.

This flexibility is important for grid operations. Thermal power plants often need time to increase and decrease output, and frequent, fast changes in generation can damage equipment or impose additional costs. Hydro and gas-based generation are better suited to respond quickly to changes in net load, and their greater use on 9 September reflects this role. Storage systems like pumped hydro plants and battery storage can also provide flexible, fast response, and play an important role in shifting generation across time, smoothing the net load associated with increasing renewable generation (Sanyal & Dev 2026). While storage was used on both 9 and 27 September, charging during solar hours and discharging in non-solar hours, its contribution was relatively small.

Figure 3: Source-wise generation at 15-minute intervals, on 9 and 27 September

Capacity utilisation

Figure 4 plots the maximum daily output of thermal, nuclear and hydro plants as a share of the generation capacity that was available to operate (installed capacity minus the capacity under outage). Hydro output remained high throughout the month, exceeding 90% and nearing 95% on several days. The high utilisation may partly reflect lower hydro availability this year due to weaker rainfall. Thermal generation also operated at close to 90% of available capacity. It is worth noting that daily aggregates hide real-time variations within a day. Maximum daily output need not coincide with the national solar or non-solar peak, and available capacity may not be fully dispatched because some generation is retained as operating headroom or reserves. Nonetheless, high utilisation rates suggest that there may be relatively little headroom to respond to unexpected increases in demand or sudden outages elsewhere in the system (Spencer and Vautrin 2026).

Figure 4: Capacity utilisation (maximum power output as a percentage of capacity on line)

Curtailment of renewable energy

On the other side of the supply story is the curtailment of solar and wind generation. Curtailment can occur for different reasons. Transmission constraints, grid-security requirements and other technical constraints may restrict the amount of renewable power being evacuated. Curtailment can also occur when there is surplus generation, particularly when demand is low or when thermal generators cannot reduce their output sufficiently. In such scenarios, the system operator can use the TRAS Down mechanism under the ancillary services framework to ask renewable generators to reduce their output even when they are technically able to generate more. This means available renewable energy is not fully utilised. In the first quarter of 2026, 300 GWh of solar and wind generation was curtailed. Better transmission infrastructure and deployment of storage systems can reduce such curtailment by allowing renewable energy to be evacuated or shifting surplus generation to periods when it is needed.

Figure 5 shows daily wind and solar curtailment for September, distinguishing between curtailment due to technical constraints and curtailment through TRAS Down. As electricity requirements eased by the end of the month, TRAS Down curtailment increased, reaching 93.58 GWh on 26 and 82.6 GWh on 27 September.

Figure 5: Solar and wind curtailment

Role of power exchanges

When the long-term contracted capacity of distribution companies is insufficient to meet demand, power exchanges provide a platform for short-term procurement. There are three power exchanges in India, but most of the trade happens on the India Energy Exchange (IEX). In our analysis, we focus on three of IEX's market segments that reflect procurement under different system conditions; Table 1 summarises their key features.

Table 1: Comparison of DAM, RTM and HP-DAM

	Day Ahead Market (DAM)	Real-Time Market (RTM)	High Price Day Ahead Market (HP-DAM)
Purpose	Procurement planned one day in advance	Balance demand and supply close to real time	Additional day-ahead procurement during periods of expected scarcity
Bids	One day before delivery	Every 15 minutes for delivery one hour ahead	One day before delivery
Activation	Used routinely for scheduled short-term procurement	Used to manage unexpected changes in demand and generation	Activated when expected prices exceed the DAM price ceiling
Price discovery	Market-clearing price at the point of demand and supply intersection, capped at Rs 10/kWh	Market-clearing price at the point of demand and supply intersection, capped at Rs 10/kWh	Market-clearing price at the point of demand and supply intersection, capped at Rs 20/kWh

In September, both the DAM and RTM were heavily stressed. Purchase bids were high in both markets, while sell bids and, consequently, market clearing volumes, were significantly lower. In the DAM, purchase bids were, on average, four times the volume of sell bids, reaching a maximum of eight times the sell bid on 17 September. In the RTM, purchase bids were, on average, five times sell bids. It was the highest at 13 times the sell bid on 16 September. The gap narrowed towards the end of the month as demand pressures eased (See Figures 6 and 7).

Figure 6: Purchase and sell bids, and market-clearing volumes in the IEX DAM

Figure 7: Purchase and sell bids, and market-clearing volumes in the IEX RTM

The divergence between purchase and sell bids can be understood by looking at both the available supply on the exchange and the market-clearing prices. Since much of India's generation capacity is tied up in long-term contracts, the amount of generation available for sale on the exchange can be limited, particularly during days of high demand. At the same time, the DAM and RTM operate under a CERC-mandated price ceiling of Rs 10/kWh. Figures 6 and 7 show the number of 15-minute blocks in each day during which the prices reached this ceiling. On most days, prices hit the ceiling in more than 60 of the 96 blocks. On 17 September, when the DAM recorded its highest purchase-to-sell bid ratio, the market clearing price reached the ceiling in 75 blocks. Similarly, in the RTM, prices reached the ceiling in 72 blocks on 16 September when the purchase-to-sell bid ratio was at its highest.

Figures 8 and 9 plot the purchase bids, sell bids, market clearing volumes and market clearing prices for each 15-minute block on 9 and 27 September in the DAM and RTM, respectively. The contrast between the two days illustrates how market conditions changed with changes in electricity requirements. On 9 September, the day with the highest electricity requirement, purchase bids were substantially higher than sell bids. Supply increased during the day, bringing down market clearing prices (MCP): in the DAM, the MCP fell to around Rs 2/kWh at 1 pm, while the RTM saw the lowest MCP of Rs 2.6/kWh at the same time block. But as solar generation declined and less supply was available, prices rose sharply and repeatedly hit the Rs 10/kWh ceiling.

On 27 September when electricity requirement was lower, market clearing volumes followed the quantum of purchase bids more closely and prices were considerably lower. The average MCP was Rs 2.76/kWh in the DAM and Rs 4.05/kWh in the RTM.

The difference in available short-term supply is also visible when comparing the two days: sell bids were consistently lower on 9 September than on 27 September for comparable time periods.

Taken together, these patterns suggest that both factors played a role during the period of highest demand. First, the price ceiling likely limited the market's ability to attract additional supply, and second, the small volume of sell bids indicates that spare generation capacity available for short-term trading may itself have been limited.

Figure 8: 15-minute blockwise buy and sell bids, MCV and MCP in the DAM

Figure 9: 15-minute blockwise buy and sell bids, MCV and MCP in the RTM

Another market in the IEX is the high price day ahead market (HP-DAM), which allows trades at prices above the ceiling price. The market was active from 4 till 17 September but traded volumes remained small, averaging around a total of 120 MWh per day. Unlike the DAM and RTM, it had higher sell bids relative to purchase bids, but on days that the market was active, HP DAM also hit its regulatory price ceiling of Rs 20/kWh (Figure 10).

Figure 10: Purchase and sell bids, and market-clearing volumes (MCV) in the IEX HP-DAM

Figure 11 plots market conditions in 15-minute blocks on 9 and 27 September. On 9 September, the HP-DAM was active in just 16 of the 96 blocks. Prices hit the Rs 20/kWh ceiling in all 16 blocks, with an average of only 9.5 MW clearing in each block. On 27 September, the market was not active, with no purchase or sell bids.

Figure 11: 15-minute blockwise buy and sell bids, MCV and MCP in the HP-DAM

Conclusion

India's electricity system is entering a period in which both demand and renewable generation are increasing rapidly. Our analysis of September 2026 shows that the power system was strained in four connected ways:

1. Shortages emerged on most days, and were concentrated during non-solar high-demand periods, when solar generation was low and the system had to rely more heavily on other sources. On 9 September, the peak shortage was 7.72 GW at around 7 pm.
2. Thermal and hydro plants ran near or above 90% of available capacity. This suggests little room to absorb demand spikes or outages.
3. The system faced both scarcity and surplus within the same month, as there was high curtailment of renewables through TRAS Down.
4. Prices on the IEX were lower when solar generation was abundant, and rose sharply as solar output declined and demand stayed high or increased. Prices hit the Rs 10/kWh ceiling in more than 60 of 96 blocks on most days.

As demand peaks rise and the share of renewable generation grows, ensuring reliability will increasingly depend not just on how much capacity is installed, but on when and where electricity is available, how quickly supply can respond, and whether it can be delivered to consumers. September's experience highlights the importance of examining these dimensions together when assessing India's electricity system.

References

Sanyal, Sanjeev and Dev, Satvik. 2026. The Duck and The Camel: Tracing the Net Load on the Indian Power Grid. EAC Working Paper Series EAC-PM/WP/53/2026.

Spencer, Thomas and Vautrin, Anthony. 2026. India's electricity demand grows at night: Managing rising cooling demand. International Energy Agency.

The code used to extract the monthly data and generate the figures in this blog is available on our GitHub repository.

The authors are researchers at TrustBridge Rule of Law Foundation. We thank Aryan Anand and Akshat Shukla for their assistance with the initial data extraction, and two anonymous referees for their helpful comments.