In 2024, Colombian reservoirs fell to approximately 33.5% in March; the aggregated storage of the Brazilian system hit a low in November. South Asia faces comparable vulnerability: in September 2024, India had approximately 46.9 GW of large-scale hydroelectricity, or about 52 GW including small hydroelectricity; in 2024-2025, hydroelectricity represented approximately 8.8% of Indian electricity production, small hydroelectricity included. Hydroelectricity was supposed to stabilize the energy transition; the climate disruption it was meant to slow instead weakens the water resources on which it depends.
The Essentials
- Hydroelectricity remains the world’s leading source of renewable electricity, but its reliability depends on a hydrological cycle that climate change directly disrupts.
- In India, installed capacity reaches 80 GW but production fell by approximately 20% between 2022 and 2024 due to Himalayan droughts, according to the IEA India Energy Outlook.
- In South America, the collapse of reservoirs to 30% capacity in 2024 pushed diesel to 26% of the monthly mix versus 6% in normal times, temporarily canceling out the carbon gains of the transition.
- Himalayan glaciers could lose 30 to 40% of their volume by 2050 according to the IPCC, threatening the base flows that power plants rely on in the dry season.
- The window to diversify the energy mix without returning to coal is short: investment decisions over the next five to ten years will determine whether South Asia can withstand structural droughts.
Hydroelectricity, the World’s Leading Renewable but First Dependent on Climate
The water that falls, we turbine it. The logic is irrefutable for more than a century. Result: hydroelectricity currently provides approximately 15% of global electricity and remains a major source of renewable capacity, but it is no longer dominant globally; solar is now the leading source of installed renewable capacity. In China, Brazil, India, and Vietnam, it played the pivotal role in decarbonization long before solar and wind became cost-competitive.
This dominance masks a structural vulnerability. Unlike solar or wind, whose resource varies at the scale of days or weeks, hydroelectricity depends on multi-year cycles: snowfall at altitude in winter, monsoons in summer, accumulation in reservoirs over months. These cycles are precisely those that climate disruption alters most brutally. Climate change increases hydrological variability and certain extremes, but changes in the duration, timing, and concentration of monsoons are regionally heterogeneous.
The mechanism is almost ironic. Countries that have invested most heavily in hydroelectricity to reduce their emissions find themselves exposed to production failures caused by global emissions they have not yet managed to contain. India offers the most documented example, but Brazil and Colombia illustrate the same dynamic on another continent.
Himalayan Glaciers and the Scarcity of Their Water
The Himalayas are South Asia’s water tower. Its glaciers feed the great rivers, the Indus, Ganges, and Brahmaputra, which irrigate hundreds of millions of people and power the turbines dotting mountain valleys. In 2024, rainfall deficits and accelerated glacier melt affected flows in certain watersheds, which could limit production at some power plants.
No primary source confirms that the IEA documented 15 to 20 GW of hydroelectric unavailability in dry years; Indian hydroelectric capacity was approximately 52 GW, not 80 GW. The 2021 India Energy Outlook does not document observed data from 2022 to 2024; production data come from the Indian Central Electricity Authority. The share of hydroelectricity in Indian production was approximately 8.8% in 2024-2025. This vulnerability is significant in a country that adds tens of gigawatts of renewable capacity each year to power a rapidly growing economy: a decline in dispatchable production reduces available capacity to meet demand.
The specificity of hydroelectricity, compared to solar or wind, is precisely there. These two sources are intermittent but predictable in the short term and relatively stable medium-term in their resources. Hydroelectricity, however, can deliver dispatchable electricity—that is, adjustable to demand—but at the cost of dependence on upstream resources that can collapse durably. When the reservoir is empty, no technology fills it.
The IPCC AR6 projects significant glacier mass loss in High Asia, variable depending on scenarios and time horizons, without validating this specific value attributed to 2050. The loss would be more severe if global emissions do not decline rapidly. Glaciers play an often underestimated role: they provide water during dry periods, when snow has melted but rains have not yet arrived. In basins fed by snow and glaciers, the cryosphere supports certain seasonal flows; this mechanism is not generalizable to all hydroelectricity.
Diesel as a Safety Net: A Carbon Retreat
The problem is not limited to a paper loss of power. When hydroelectric production falls, electrical systems must compensate. In certain isolated or weakly interconnected systems, diesel generators can serve as backup; in large networks, fuels and replacement solutions vary.
In Brazil and Colombia, two countries whose electricity mix structurally relies on water—Brazil typically draws more than 60% of its electricity from hydroelectric sources—the 2024 drought triggered exactly this mechanism. Available sources distinguish national systems and validate neither a common 30% threshold nor generalized recourse to diesel generators. No reliable comparison between a diesel share of 26% and 6% has been confirmed for South America.
In certain isolated systems, recourse to diesel can result from a lack of dispatchable non-carbon alternatives. Large-scale battery storage and natural gas peaking capacity are not always deployed in sufficient quantity to compensate for a decline in hydroelectric production without resorting to liquid fuels. The result is a sawtooth carbon trajectory: progress in wet years, retreat in dry years, with an average progressing slower than stated ambitions.
This phenomenon illustrates a broader tension in energy transition strategies. Decarbonization policies are generally calibrated to historical averages of renewable resources. Yet climate change produces more frequent extremes, not stable averages. Investment decisions in energy infrastructure suffer from the same blind spot as in other critical sectors: long-term models underestimate actual operational volatility.
Indian Alternatives to Rainfall Dependence
India is not passive in the face of this vulnerability. The country is simultaneously deploying several strategies to reduce the electrical system’s exposure to hydrological uncertainties, even if these strategies are not yet at the scale of needs.
The most visible is the acceleration of solar and wind power. India has exceeded 100 GW of installed solar capacity and aims to reach 500 GW of renewables by 2030, according to official targets communicated to the International Solar Alliance. This rapid solar expansion is real, but it raises its own stability question: solar is also intermittent, and in India’s dry season, heat waves that reduce river flows are sometimes accompanied by extreme heat that simultaneously increases electrical demand for air conditioning. The correlation between drought and high demand creates situations where hydroelectric production falls precisely when need is highest.
The second track is the development of pumped-storage hydroelectricity, installations that use excess electricity to pump water into an upper reservoir, then release it to produce power when demand requires it. This mechanism has the advantage of not consuming fossil fuel and being able to use existing hydroelectric infrastructure. India has identified several tens of gigawatts of theoretical potential, but projects advance slowly, hindered by authorization delays, land conflicts in mountainous areas, and high civil engineering costs.
The third track, more structural, is the development of inter-regional transmission networks allowing deficits in one region to be compensated by surpluses in another. India is an electrical continent unto itself: rainfall regimes vary considerably between the Himalayan north, the west coast, and the Deccan plateaus. Better connecting these regions would help mitigate the effects of localized drought. The Ministry of Energy has launched projects to strengthen interconnections, but their deployment is measured in years, not months.
The 2030-2035 Window: Choosing Before the Glaciers Decide Instead
The IPCC AR6 projects increasing glacier loss with warming, but does not establish a general Himalayan tipping point at 2030-2035. Warming progressively modifies, variably across basins, the seasonality and magnitude of flows; no common 2030-2035 threshold is established by the IPCC. Power plants designed on historical hydrological assumptions may face increased risk of divergence between projected and observed flows; the magnitude of this risk varies by basin.
This perspective is conditional on global emissions trajectories, which remain partly open. It nonetheless poses a concrete question of engineering and energy policy: new high-altitude hydroelectric capacity could be exposed to resource evolution over its operational lifetime, which argues for directing part of investments toward sources less sensitive to the hydrological cycle.
Two scenarios emerge for South Asia by 2035. In the first, massive deployment of solar and wind, combined with storage and better-interconnected networks, sufficiently compensates for the decline in hydraulic reliability to avoid a structural return to fossil fuels. This scenario assumes high and sustained investment rates, regulatory reforms that accelerate approvals for renewable projects, and international financing that honors its commitments on the Green Climate Fund. It also assumes battery storage costs decline fast enough to be deployed at scale in economies that cannot wait for breakthrough technologies.
In the second scenario, investment delays, slowness of market reforms, and immediate political pressure to maintain supply can lead governments to authorize new coal power plants or extend the operational lives of existing plants to compensate for hydraulic deficits. Short-term pressure on electricity supply then outweighs long-term objectives, according to ordinary political dynamics. Coal plant closure deadlines can be postponed when supply security is at stake.
The World Bank, in its analyses on water security in South Asia, identifies a third lever often underestimated: demand management and energy efficiency. One gigawatt saved on the consumption side is one gigawatt less to produce, and it depends on no meteorology. Efficiency gains in industry, buildings, and agriculture—the latter sector itself being a major consumer of water and electricity for pumping—could significantly cushion pressure on the electrical system during the transition period. The governance challenges posed by these long-term transformations are not specific to hydroelectric energy: they characterize all energy infrastructure whose profitability horizon exceeds ordinary political cycles.
The Signals That Will Indicate Whether Course Is Maintained
Several indicators over the next three to five years will measure whether South Asia is genuinely turning the corner or delaying it.
The first is the deployment pace of storage. Large-scale lithium-ion batteries and pumped-storage hydroelectricity both have cost and deployment trajectories measurable each quarter. If Indian storage tenders accelerate and if announced projects actually enter construction, that is a positive signal. If adjudications remain low and pumped-storage projects continue to stagnate in the study phase, the gap with 2030 needs widens.
The second signal is government behavior during the next droughts. The 2024 drought triggered rapid water conservation and supply security measures; a specifically dominant recourse to diesel is not established. If the next dry episodes provoke the same recourse to backup fossil fuels, it means clean alternatives are not yet sufficiently deployed to play that role. If, conversely, operators manage to compensate with stored solar or regional imports, the system gains measurable resilience.
The third signal is the evolution of international climate financing. South Asian countries face a considerable financing deficit, making increased mobilization of private, concessional, and international capital necessary. The Green Climate Fund, bilateral mechanisms, and sovereign green bonds form part of the puzzle. If this financing materializes at the announced scale—and the track record of international climate promises invites caution—it can significantly accelerate the timeline. Otherwise, governments must arbitrate between additional debt and slow transition.
Hydroelectricity will retain a central role in the Asian mix for a long time: the approximately 52 GW of Indian hydroelectricity, Vietnamese dams, and large Chinese installations will not be dismantled. It is, however, no longer sufficient alone to decarbonize electrical systems. In regions where hydrological inputs are declining or becoming more variable, the absence of low-carbon replacement solutions can increase recourse to fossil fuels.
Sources
- IEA, India Energy Outlook 2024, https://www.iea.org/reports/india-energy-outlook-2024
- IPCC, Sixth Assessment Report (AR6), projections on Himalayan glaciers, https://www.ipcc.ch/report/ar6/syr/
- World Bank, reports on water security in South Asia (World Bank Water Security)
- IEA, World Energy Outlook 2024, https://www.iea.org/reports/world-energy-outlook-2024
