Maharashtra needs electricity. There is little argument about that. As industries expand, cities grow and electricity consumption increases, the state needs a reliable supply throughout the day. At the same time, solar and wind energy are becoming increasingly important. But the sun does not shine throughout the night, and wind generation can change with the weather.
That creates a genuine electricity-management challenge. One possible solution is pumped-storage hydropower. It allows electricity to be stored by pumping water uphill and recovered later by releasing it through turbines.
The technology is not new. It is used in several countries, including India. It can help manage electricity demand, balance renewable generation and provide power during peak hours.
But there is a question Maharashtra must answer before citizens are expected to understand the environmental and financial consequences of a major project: when a project is described as clean energy, what exactly are we counting as clean, and what are we counting as its cost?
That question has become particularly important around Sai Dongar in Karjat, Raigad district. The proposed Saidongar-1 pumped-storage project, developed by the Torrent Power group, has an installed capacity of 3,000 MW. Project records put its estimated cost at approximately Rs 13,505 crore. The project documents identify a land requirement of around 377 hectares, including approximately 233 hectares of forest land.
A separate 1,200 MW Saidongar-2 proposal is also listed by the Maharashtra Pollution Control Board. If both projects proceed as proposed, their combined installed capacity would be 4,200 MW. Their shared facilities and cumulative impacts therefore deserve careful public examination.
This is not an argument against pumped storage. It is an examination of whether the technical, financial, environmental and public-interest case for these projects has been explained sufficiently.
1. What exactly is Maharashtra building near Karjat?
The project is officially identified as Saidongar-1 – Karjat Open Loop Pumped Storage Project. The Maharashtra Pollution Control Board lists it as a 3,000 MW project proposed by Torrent PSH3 Private Limited. Its public hearing was conducted on 3 September 2025.
According to the project documents, the development requires approximately 377 hectares of land: 233 hectares of forest land and 144 hectares of non-forest land. The proposed infrastructure includes upper and lower reservoirs, underground and associated hydro-mechanical works, access roads and transmission infrastructure.
Recommended Product
Amazon Renewed — Best Deals of the Day
🛒 View on Amazon →As an Amazon Associate, we earn from qualifying purchases. Price and availability may vary.
The project is designed as an open-loop pumped-storage facility involving the Pej River system. An open-loop project has a connection with an existing water body or river system. A closed-loop project is designed to operate primarily between reservoirs without the same continuing river connection. This makes the project’s water assessment particularly important.
The Ministry’s appraisal records identify the project as a 3,000 MW open-loop scheme. The developer’s compliance report records environmental clearance dated 12 January 2026. Forest clearance is a separate statutory process, with its own stages and conditions. These approvals should not be confused with one another.
Our question to the government: Can the complete project approval record, including every environmental and forest-clearance condition, be made easily accessible to the public in one place?
Sources: MPCB – Saidongar-1 project record · Torrent Power – June 2026 environmental compliance report.
2. The 3,000 MW headline does not tell us how much electricity is stored
This is perhaps the most important technical distinction in the discussion. MW measures power capacity. MWh measures energy.
A plant rated at 3,000 MW does not necessarily mean it can supply 3,000 MW continuously for an entire day. The draft environmental assessment describes approximately six hours of generation at full rated output.
The arithmetic is straightforward: 3,000 MW multiplied by six hours equals 18,000 MWh, or 18 GWh. In simple terms, the plant could deliver 3,000 MW for six hours under the stated operating assumptions.
That is a substantial amount of power. However, the storage duration is just as important as the headline capacity. A project that can deliver 3,000 MW for six hours has a different energy capability from one that can deliver the same power for ten hours.
The public should therefore be given the complete technical specifications, including usable storage volume, operating water levels, pumping requirements and guaranteed generating duration.
Our question to Torrent Power: Why should the public be expected to evaluate a 3,000 MW project without a clear explanation of its usable MWh capacity and operating cycle?
3. Pumped storage does not create electricity. It stores electricity.
Let us understand the technology without getting lost in engineering terminology. During periods when electricity is available and the plant is scheduled to pump, electricity runs pumps that move water from a lower reservoir to an upper reservoir. Later, when demand increases, water is released through turbines. The turbines drive generators and electricity is supplied to the grid.
The water is reused through the operating cycle, subject to losses and replenishment requirements. However, the electricity returned is lower than the electricity consumed in pumping. This is called round-trip efficiency.
For example, assume a pumped-storage plant consumes 100 units of electricity while pumping water uphill. If its round-trip efficiency is 75%, it returns approximately 75 units. The remaining 25 units represent energy losses.
This does not make pumped storage useless. Electricity available during a period of surplus may be more valuable when demand is high. But the source of the pumping electricity matters.
If the electricity used for pumping comes from renewable generation that would otherwise have been curtailed, the project may help use energy that would otherwise go unused. If it comes from coal-based generation, the calculation becomes more complicated. The system may still provide grid services, but its net emissions benefit must be assessed using the actual electricity mix, operating schedule and generation displaced.
Our question to the government and Torrent Power: What is the expected source of pumping electricity, and what arrangements will ensure that the project uses surplus low-carbon electricity when it is available?
4. Where will the pumping electricity come from?
This is not a theoretical question. It is central to the project’s environmental justification.
Maharashtra has a substantial renewable-energy installation base. However, installed capacity is not the same as electricity actually generated. Solar generation is concentrated during daylight hours. Wind generation varies with weather. Electricity demand also changes throughout the day.
A pumped-storage plant needs electricity at the times when pumping is scheduled, and the grid must be able to deliver that electricity to the plant.
The MSLDC monthly generation report for May 2026 provides one illustration. It reports 10,677.27 million units of thermal generation, 1,068.39 million units of solar generation and 898.25 million units of wind generation, among other sources. Solar and wind together accounted for approximately 14.3% of the reported in-state generation in that particular month.
This is not an annual renewable-energy share, and it does not represent the complete picture of electricity supplied to Maharashtra consumers. The state also exchanges electricity with other regions and makes purchases. Nor does a monthly total tell us how much renewable electricity was available for pumping at any particular hour.
It does, however, show why annual installed-capacity figures cannot settle the question.
Our questions to MSLDC and the Energy Department:
- How many hours in a typical year does Maharashtra have surplus solar or wind electricity?
- How much renewable electricity is curtailed?
- At what times would Saidongar-1 normally pump water?
- What is the expected generation mix during those pumping hours?
- Is the project expected to use dedicated renewable-energy contracts?
- How will the state measure the project’s actual carbon benefit?
The answers should be supported by hourly grid data, not simply annual renewable installation figures.
Source: Maharashtra State Load Despatch Centre.
5. What happens to the forest?
The forest question has become the most visible part of the controversy. According to the project documents, approximately 233 hectares of forest land are involved. Reports in October 2026 referred to permission for the felling of 32,490 trees, followed by a reported pause after protests.
The tree figure should be read carefully. Permission to fell trees does not establish that all those trees have been cut. The latest written status of the permission and any subsequent stop instruction should be made public.
There is also a larger ecological question. A forest is not simply a collection of individual trees. It is a living system involving vegetation, soil, insects, birds, mammals, fungi, streams and communities that depend on it. When an established forest is cleared, the consequences cannot necessarily be measured by counting saplings planted somewhere else.
The developer has referred to compensatory afforestation and additional plantation. Torrent has also said that the project is outside the Bhimashankar Wildlife Sanctuary and its notified Eco-Sensitive Zone. These statements should be examined alongside the actual clearance maps, ecological studies and forest records.
The Ministry’s forest advisory records describe a change in the originally proposed compensatory-afforestation land and consideration of an alternative site in Satara, subject to conditions. Citizens deserve to know what ecological value the replacement land has, whether it can support the proposed plantation, and how survival will be monitored over the coming decades.
The tree-felling pause
News reports on 28 October 2026 said that the proposed tree-felling operation had been halted following protests. This is an important distinction: the permission was reported as having been granted, while the subsequent operation was reported as paused. A pause does not automatically cancel the permission or the project.
Our questions to the Forest Department:
- What is the exact forest area being diverted?
- What is the complete tree inventory?
- How many trees have actually been felled?
- What is the present written status of the October permission?
- What conditions govern any future felling?
- Where is the compensatory-afforestation land?
- How will survival and ecological restoration be independently audited?
Sources: India Today – Karjat tree-felling controversy · The Times of India – reported pause in tree-felling plan.
6. The water question: what happens to the Pej River?
The project is an open-loop pumped-storage facility associated with the Pej River system. The Ministry’s appraisal documents give an initial filling requirement of approximately 23 million cubic metres and an additional annual water requirement of approximately 3 million cubic metres.
The developer’s position is that the water requirement can be met through the identified catchment and that downstream irrigation and drinking-water users will not be affected. That is an important assurance, but it should be examined against the underlying hydrological study.
Water availability in a river catchment is not constant throughout the year. Rainfall varies, monsoons can be weak and dry periods can last longer than expected. Local communities and ecosystems depend on water that may not appear in a project’s financial calculations.
The public needs access to the complete hydrological study. It should explain the source and timing of initial filling, seasonal water availability, assumptions used for dependable water yield, evaporation and seepage estimates, annual replenishment requirements, environmental-flow provisions and downstream water availability.
The developer should also explain whether its assessment considers the combined implications of Saidongar-1 and Saidongar-2.
Our question: What happens if actual water availability is lower than the assumptions used during project appraisal?
7. The construction footprint is much larger than a photograph of a reservoir
There is another part of the project that deserves public attention: excavation.
The Saidongar-1 draft environmental assessment estimates approximately 110.78 lakh cubic metres of excavated material. The project documents identify proposed muck-disposal locations covering approximately 41 hectares.
That is a substantial volume. The public needs to understand where this material will be taken, how it will be transported and how disposal locations will be protected.
Excavation and disposal can affect natural drainage, hill slopes, streams, soil stability, forest access, local roads, construction traffic, dust and noise levels. There are also tunnels, access roads, powerhouses, transmission infrastructure and construction facilities to consider.
The total environmental footprint cannot be understood by looking only at the upper and lower reservoirs.
Our question to the developer and government: Can the complete construction footprint, including every road, tunnel, muck-disposal site and transmission corridor, be published on an accessible map?
8. What is the financial justification?
The project has an estimated cost of approximately Rs 13,505 crore in the Ministry’s appraisal record. The draft environmental assessment cited a different earlier estimate of approximately Rs 13,017 crore. The latest approved cost and the reason for the change should be explained rather than silently treating the two figures as identical.
The developer’s appraisal submissions include projected annual electricity generation of approximately 6,241.5 million units and long-term projections of electricity-sale benefits and economic contribution. These are projections, not guaranteed returns.
They depend on assumptions about electricity prices, pumping costs, financing, operating expenditure, maintenance and the number of operating cycles. A pumped-storage project may have a long operating life, but that does not automatically guarantee a good investment.
If 100 units are consumed to pump water and 75 units are returned, the project must recover the cost of the 100 units consumed while earning revenue from the 75 units returned, together with any capacity or grid-service payments. That is why the commercial structure matters.
Our questions to Torrent Power and the regulators:
- What is the expected annual pumping-energy cost?
- What electricity prices are assumed?
- What is the expected annual number of operating cycles?
- What round-trip efficiency is guaranteed?
- What revenue will come from energy sales?
- Will there be capacity payments or ancillary-service payments?
- Who bears the risk if the plant is used less frequently than expected?
- How will the project’s costs affect electricity consumers?
- What is the financial model’s sensitivity to lower utilisation and changing market prices?
The public does not necessarily need every commercially sensitive detail. But the assumptions used to justify a major infrastructure investment should be transparent enough for independent scrutiny.
9. What can India learn from the Tehri pumped-storage project?
Pumped storage is not an untested idea in India. The Tehri Pumped Storage Plant in Uttarakhand provides a useful domestic comparison.
According to THDC India, the facility has 1,000 MW installed capacity, four reversible units of 250 MW each, the Tehri reservoir as its upper reservoir and the Koteshwar reservoir as its lower reservoir. THDC publishes a design round-trip efficiency of approximately 79.41%.
Tehri is a useful comparison, but its physical setting differs from Sai Dongar because it uses reservoirs associated with existing dams. Its environmental footprint and financial model cannot simply be transferred to a project requiring new reservoirs and associated land and forest diversion.
The lesson is not that every project will perform the same way. It is that actual operating data, efficiency, grid conditions and the source of pumping electricity matter.
The question for Sai Dongar is not whether pumped storage can work. It is whether its own design, location, water availability, electricity source and financial model justify its proposed scale.
Source: THDC India – Tehri Pumped Storage Plant.
10. What can China teach us? The Fengning example
China’s Fengning Pumped Storage Power Station is another useful comparison. The facility has an installed capacity of 3,600 MW and entered full operation in December 2024.
Its published design figures include annual electricity generation of 6.612 billion units and annual pumping consumption of 8.716 billion units. Those figures imply an annual design energy ratio of approximately 75.8%.
The station is located near the Beijing-Tianjin-Hebei load centre and a large renewable-energy development area in northern Hebei. Published accounts describe its role in absorbing renewable electricity and supporting the regional grid.
This is an important lesson. Pumped storage becomes particularly useful when it is connected to a grid that has renewable electricity available at the right times and a demand centre that needs electricity later.
Fengning’s experience does not automatically prove that every proposed pumped-storage project in Maharashtra will deliver the same benefits. Its design figures are not a performance prediction for Sai Dongar.
Can Maharashtra demonstrate the same relationship between renewable-energy surplus, storage operation and electricity demand for Saidongar-1?
Source: Fengning Pumped Storage Power Station – published project information.
11. Who is actually building Saidongar-1?
It is important to distinguish the project developer from the engineering contractors.
Torrent Energy Storage Solutions Private Limited, a Torrent Power group company, is the project developer.
ANDRITZ has announced a contract to supply reversible pump-turbine units, motor-generators and associated electromechanical equipment. Its announcement gives an annual generation projection of approximately 6,241 GWh and estimates annual carbon-dioxide reductions of 4.55 million tonnes. These are company projections, and the assumptions behind them deserve examination.
Larsen & Toubro announced a contract from Torrent Energy Storage Solutions in January 2026 for civil and hydro-mechanical works associated with the project.
These are substantial responsibilities. However, the developer, equipment supplier and civil contractor have different roles. The developer remains responsible for the overall project and its commitments. Contractors are responsible for their respective contractual scopes.
There is also a specification that deserves clarification. The October 2025 appraisal record lists nine 300 MW units and two 150 MW units. The later contractor announcements describe ten 300 MW units. Both configurations total 3,000 MW, but the difference should be explained by the developer and contractors.
Questions for the companies: What comparable projects have they delivered? How did those projects perform against their original cost and commissioning schedules? What operating efficiencies were achieved? What lessons have been incorporated into Saidongar-1? What performance guarantees apply to their respective contracts?
Sources: ANDRITZ – Saidongar equipment contract · L&T – Saidongar civil works contract.
12. There is another project: Saidongar-2
The investigation becomes more important when we look beyond Saidongar-1.
The Maharashtra Pollution Control Board separately lists a proposed 1,200 MW Saidongar-2 Maval Pumped Storage Project, associated with Torrent PSH4 Private Limited and covering locations in Karjat taluka, Raigad district, and Maval taluka, Pune district.
If both projects proceed as proposed, their combined installed capacity would be 4,200 MW. Their combined environmental footprint should be examined, not assumed.
The draft environmental documents indicate that Saidongar-1 and Saidongar-2 are linked through a common lower-reservoir arrangement and a shared catchment. The Saidongar-2 draft identifies 1,200 MW capacity, six hours of generation, approximately 141.44 hectares of land including 35.6 hectares of forest, and a separate excavation and muck-disposal requirement.
It also refers to catchment-treatment planning being included in Saidongar-1’s environmental plan. The developer and authorities should explain how these shared environmental obligations will be funded, implemented and monitored across both projects.
Our questions:
- What facilities do the two projects share?
- Have their combined water requirements been assessed?
- Have cumulative forest and biodiversity impacts been examined?
- How will shared catchment-treatment commitments be implemented?
- Have the combined effects on local villages, roads and transmission infrastructure been assessed?
Source: MPCB – Saidongar-2 project record.
13. Maharashtra’s proposed 81,000 MW pumped-storage programme: how much is actually needed?
The Sai Dongar project is part of a much wider state push for pumped-storage development. Government announcements have referred to approximately 59–60 proposed projects, with a combined capacity of around 81,000–82,000 MW and an estimated investment of approximately Rs 4.28 lakh crore.
These figures must be understood correctly. They refer to a proposed project pipeline and associated investment estimates, not to commissioned storage capacity. A memorandum of understanding is not a completed project. A proposed project is not an approved project. An approved project is not necessarily under construction.
The Central Electricity Authority’s national records provide useful context. A Government of India update in August 2026 reported 7.43 GW of operational pumped-storage capacity, 15.87 GW under construction or implementation, and 10.40 GW in projects concurred with or appraised by the CEA.
The scale of Maharashtra’s announced pipeline therefore deserves a transparent project-by-project explanation. What is the expected electricity demand? How much renewable generation is projected? What storage duration is required? How much of the announced capacity is likely to be commissioned, and when?
Our question to the Maharashtra government: What technical study demonstrates that the state needs the proposed scale of pumped-storage capacity, and how was the combined figure calculated?
Sources: Central Electricity Authority – Pumped Storage Plants · Press Information Bureau – India’s pumped-storage pipeline, August 2026.
14. Are there alternatives, and have they been compared properly?
Pumped storage is one option. It is not the only option. Other technologies and operating methods can help balance electricity supply and demand.
Battery storage
Grid-scale batteries can respond quickly and are increasingly being deployed alongside solar and wind generation. However, batteries have their own capital costs, replacement requirements, material supply chains and duration limitations.
Thermal power flexibility
Coal-based power plants can sometimes reduce their output when renewable generation is high and increase it when demand rises. However, flexible operation can increase operating stress, maintenance costs and fuel consumption per unit of electricity at lower loads. It also cannot provide every service that long-duration storage can provide.
Transmission improvements
A stronger transmission network can allow electricity to move from areas with surplus generation to areas where demand is higher. However, transmission infrastructure also requires land, investment and environmental planning.
Demand response
Some industrial and commercial consumers can shift electricity consumption to different hours. This can reduce peak demand and help the grid operate more efficiently.
The real question may not be whether Maharashtra should choose batteries, pumped storage or thermal flexibility. It may be how much of each is required, where it should be located, and what combination provides the required reliability at a reasonable overall cost.
The state should publish the alternatives analysis used to justify the proposed pumped-storage pipeline.
15. What about the people who live there?
Infrastructure discussions often concentrate on megawatts, investment and project timelines. However, the people living in and around a project area experience its consequences directly.
They may depend on forests for fuel, fodder and other resources. They may rely on local water sources, agriculture, grazing land and traditional access routes.
The environmental assessment should be examined alongside the actual concerns raised by affected residents. Questions about land acquisition, rehabilitation, compensation and Gram Sabha consultation should be answered with documents rather than general assurances.
The government should explain which villages are affected, which households may lose land or access to resources, and what arrangements exist for compensation and rehabilitation. The developer should publish its community-engagement and grievance-redressal arrangements.
A project may serve electricity consumers across a large region. Its local consequences, however, are experienced by a much smaller number of communities. Both sides of that equation deserve attention.
16. What we are asking the Maharashtra government and Torrent Power
This investigation is not a declaration that the project should be cancelled or that it should proceed without further discussion. It is an invitation to explain the decisions and assumptions behind it.
Questions for the Maharashtra government
- What is the state’s documented requirement for pumped-storage capacity up to 2035–36?
- What technical modelling supports the proposed 81,000–82,000 MW pipeline?
- How many announced projects have completed statutory clearances?
- How many have begun construction?
- What is the expected cumulative forest diversion?
- Has the state assessed the combined impact of multiple projects in ecologically sensitive landscapes?
- What independent assessment supports the renewable-energy surplus assumptions?
- What alternatives were considered before approving the proposed scale?
- How will consumers be protected from unnecessary financial commitments?
- Where can citizens access the complete project register and approval documents?
Questions for Torrent Power
- What is the exact usable storage capacity of Saidongar-1 in MWh or GWh?
- What is its guaranteed round-trip efficiency?
- What electricity sources will supply the pumps?
- What is the expected annual pumping consumption?
- What is the expected annual generation?
- What assumptions support the company’s carbon-reduction projections?
- What is the complete water-balance assessment?
- How will the company monitor impacts on the Pej River system?
- Where will the excavated material be disposed of?
- How will the company protect local biodiversity?
- What is the current written status of tree-felling permission?
- What are the combined implications of Saidongar-1 and Saidongar-2?
- What arrangements are in place for affected communities?
- How will environmental commitments be independently audited?
- What is the expected commercial model and who bears utilisation risk?
Questions for ANDRITZ and L&T
- What comparable projects have you delivered?
- What technical lessons are being applied at Saidongar?
- What performance guarantees apply to your respective contracts?
- What quality, safety and environmental monitoring systems will be used?
- How will the project address geological and construction risks?
- How will the public be informed about progress and compliance?
These are reasonable questions for projects of this scale.
17. The public deserves the complete picture
There are two facts that can exist together. Maharashtra needs reliable electricity, and pumped storage can help provide it. At the same time, the construction of pumped-storage projects can have significant consequences for forests, rivers, land, communities and public finances.
One fact does not cancel the other.
The experience of Tehri and Fengning demonstrates that pumped storage can serve important electricity-system needs. It also demonstrates why operating data, efficiency, grid conditions and the source of pumping electricity matter.
Saidongar-1 must be examined on its own design and location. Saidongar-2 must be considered in the wider landscape. Maharashtra’s proposed pumped-storage programme must be justified by a transparent assessment of future demand, renewable generation, alternatives and total cost.
The public should not have to choose between electricity security and environmental accountability. The two must be considered together.
A 3,000 MW project deserves more than a 3,000 MW headline. It deserves a complete explanation.
NewsPatron invites responses
This article is being published to encourage an informed public discussion about the proposed Saidongar pumped-storage projects and Maharashtra’s wider electricity-storage programme.
The questions raised here arise from publicly available project documents, government records, company announcements, technical material and reported developments. They are questions, not allegations of wrongdoing. Where figures represent company projections, government estimates or reported developments, they have been identified as such.
NewsPatron invites the Maharashtra government, the Ministry of Environment, Forest and Climate Change, Torrent Power, ANDRITZ and Larsen & Toubro to respond to the questions raised in this investigation. We also welcome technical explanations, relevant project documents, independent research and representations from residents and communities in the affected areas.
Any substantive response or verified correction received after publication will be considered for a transparent update to this report.
The purpose is not to prejudge the projects. It is to ensure that the questions surrounding them receive clear, evidence-based answers.
NewsPatron — Let Curiosity Be Your Guide.
