Is the Bhimashankar protest justified?
What 32,490 trees, an 18 GWh storage proposal, and American reservoir projects can actually tell us.
Hal Varian is the economist intended by the tag. The tag describes an analytical aspiration, not attribution or endorsement.
A justified protest. A conditional project.
My answer is yes, as a demand for accountability and a convincing demonstration that this particular project is worth its costs. The evidence here does not establish that every version should be cancelled. A protest can identify a valid concern without establishing that its preferred final outcome is correct.
My starting question was simple: have comparable projects in the United States cut similar numbers of trees? If they have, would that make opposition in Maharashtra unreasonable? Another country cutting more trees is evidence about what infrastructure can require, not evidence that this forest should be cleared. The useful question is whether the same electricity-system service can be delivered on a realistic timeline at lower total cost to society.
Here, “Bhimashankar protest” means the controversy over Torrent’s Saidongar-I–Karjat pumped-storage project in Maharashtra. It is a dispute about an actual place, a forest footprint, and who gets to bear the cost of a more reliable grid.
As of October 27, reporting says tree felling was halted on October 26. A halt is not cancellation, and the reported permission covering 32,490 trees is not a verified count of trees already felled. [1]
Three numbers, three different questions
How quickly the plant is designed to deliver electricity.
The energy associated with six hours at rated output.
The forest component of a 377-hectare project footprint.
Pumped storage moves water uphill using electricity, then releases it through turbines when electricity is more valuable. It shifts energy through time and loses some along the way. MWh here measures stored energy, not annual generation. [8] Its climate value depends on what charges it and what generation it displaces. The six-hour design is useful capacity, not a promise of uninterrupted output every day. [2] [3]
Take both sides seriously
The strongest case for protest
Forest is more than standing timber. Habitat connections, water flows, soils, mature trees and local uses cannot be replaced on the day saplings go into the ground.
The costs are concentrated around the project; electricity benefits can spread far beyond it. Potential losses of access, land or livelihoods require scrutiny of rights, consultation, safeguards and remedies, not just assurance that aggregate benefits are positive. These are questions to evaluate, not a finding that every feared outcome or violation has occurred.
Waiting for disclosure can preserve choices. Clearing first destroys that option. Protest can be a rational response to uncertainty when the harm is hard to reverse.
The strongest case for the project
A grid with more variable generation needs ways to meet demand when supply is scarce. A six-hour resource could serve evening demand after solar declines. Its actual value depends on charging mix, dispatch, transmission, reliability and cost. Bath County has operated since 1985, illustrating possible durability, not proving Karjat’s economics. [7]
A decision to reject the project should account for how the service will instead be supplied, or why it is unnecessary. Batteries require materials and replacement; other sites and transmission use land. Less storage can change generation choices or reliability. Citizens need not design an entire replacement power system before they may object.
The developer and approving authority retain the burden of demonstrating site choice and safeguards. The project has an environmental clearance with conditions. The honest defence is to demonstrate that this design meets the service requirement better than feasible alternatives, with enforceable safeguards—not to claim that storage is impact-free. [2]
These arguments are not equally strong on every question. The grid’s need for flexibility is not evidence that this exact site is best. The existence of forest loss is not evidence that every alternative has lower total harm. The missing comparison sits between those two claims.
A tree count is a clue, not a verdict
Gross Reservoir’s expansion in Colorado offers a documented example of large forest clearing for infrastructure. Bath County offers a close comparison in pumped-storage generating capacity. They answer different questions.
| Project & purpose | Capacity / land context | Tree evidence | What it can tell us |
|---|---|---|---|
| Saidongar-I–Karjat Electricity storage | 3,000 MW; 18,000 MWh 233 ha forest within 377 ha project | Reported permission 32,490 trees Verified actual-felled total: unknown | The proposed local scale; not a completed clearing total. [1] [2] |
| Gross Reservoir expansion Water supply | About 415 forest acres ≈168 ha in the tree-removal plan | Planned About 200,000 trees Actual area, not a tree count About 100 acres of initial construction trees/brush cleared in 2022 | A footprint reference. It supplies stored water, not an equivalent electricity-storage service. [6] |
| Bath County Electricity storage | 3,003 MW net generation capacity; operating since 1985 | Unknown Original tree-clearing total not established here | A generating-capacity comparator, not a verified tree-footprint comparator. [7] |
| Cabin Creek, Colorado Historical pumped storage | Smaller-scale plant | Unknown Original tree tally not established | Historical aerial work does not support a reliable tree estimate here. |
Do not add the Gross planned count to its earlier cleared acreage, subtract one from the other, or treat either as a verified completed tree total. Acre conversion uses 1 acre = 0.404686 ha; figures are rounded. Gross’s amended removal threshold is at least 2 inches diameter at breast height and 5 feet tall. Different inventory thresholds, forest types and footprint definitions preclude a defensible tree ratio or trees-per-MW ranking.
“America cut more trees” would not settle whether Karjat is justified. Nor would “Karjat has fewer trees per hectare” establish lower ecological value. Unknown is not zero. We also omit a Cabin Creek tree estimate: the historical aerial work available to us cannot support a reliable count.
Compensation is a commitment, not instant equivalence
The May 2026 forest advisory minutes discuss replacement compensatory-afforestation land of 252.7597 ha, with canopy density 0.7 and a crop-improvement plan. That is not a description of an entirely bare site becoming new forest. Its area cannot be booked as a proven net addition of equivalent habitat. [4]
The relevant questions are additional ecological improvement, survival, species composition, location, monitoring, and time. A hectare elsewhere may satisfy a specified requirement without replacing the local forest’s functions. The environmental clearance itself also says other applicable approvals remain necessary. [2]
Compare the same service, then expose the trade-offs
The economic question is whether this project is preferable to the best feasible way of meeting the same reliability need. No arbitrary “green score” or invented rupee value for a forest is needed to ask it.
- Specify the service. Which hours, delivery point, duration, reliability requirement and commissioning date must be met?
- Compare portfolios. Test alternative layouts and sites, batteries, demand response and network options against that requirement. The draft EIA considers four layouts and selects among them using environmental, social and engineering criteria. That is real analysis, but narrower than a comprehensive comparison with batteries, existing reservoirs, alternative sites, demand response or hybrid portfolios. [5]
- Show ecological differences. Publish affected habitat, water impacts, fragmentation, construction disturbance and residual harm after mitigation.
- Show who gains and who pays. Distinguish developer returns, consumer reliability, public costs and losses borne by nearby communities.
- Make promises enforceable. Tie protections to disclosed milestones, independent monitoring, remedies and consequences for non-compliance.
Conceptual net social value = electricity-system benefits − lifetime financial costs − environmental and social costs. This is a framework, not a calculated result. Use a common horizon and include charging electricity, losses, construction, maintenance, replacement, transmission, water and closure.
Ask about switching values: how much extra would a lower-impact alternative cost? What changes in construction cost, charging cost or mitigation success would change the preferred option? Where credible monetary values are unavailable, report biodiversity and local effects separately.
Some impacts can be priced; some involve rights or ecological limits that a positive total benefit cannot simply erase. A transparent comparison should show both, alongside uncertainty.
Make the judgment falsifiable
Evidence that would strengthen conditional approval
A credible comparison showing this layout supplies the required service with lower overall harm than feasible alternatives; independently checked ecological and hydrological work; documented protections for affected people; and published, enforceable compliance conditions.
Evidence that would strengthen relocation or cancellation
Unacceptable habitat or water damage, unresolved community rights, material failures in approvals or compliance, or a feasible alternative delivering the same service at substantially lower social and ecological cost.
Evidence needed before clearing resumes
A reconciled tree inventory by species, size and location, matching permission totals, measurement thresholds and actual removals; a full disturbance map separating temporary and permanent reservoirs, roads, transmission, spoil and construction; comparable alternatives with costs and dates; independent ecological, water and geotechnical review including cumulative effects; and the local-rights and implementation record with enforceable safeguards. This is a proposed evidentiary standard, not an assertion that every item is absent or a legal violation has been established.
A pause has value when it produces evidence that changes the decision. Indefinite delay without a route to a decision has costs too. The right question is whether this investment in this landscape is the best defensible way to meet a real need.
Sources you can inspect
Evidence status: primary documents establish specifications and recorded decisions; reporting establishes attributed current-status claims; actual felling totals and ecological equivalence remain unverified. No numerical ranking is claimed.
Facts dated October 27, 2026. Planned quantities, reported permissions and observed outcomes are labelled separately. This story uses public sources; it does not rely on private correspondence. Policy judgments above are the author’s interpretation.
- Times of India, October 27, 2026: reported permission for 32,490 trees and October 26 halt; not an audited count of completed felling.
- MoEF&CC environmental clearance, January 12, 2026: project identity and capacity; pp. 24–25 land, sanctuary/ESZ distances and recorded wildlife; p. 2 scope of clearance.
- EAC minutes, January 2025, PDF p. 125: 18,000 MWh storage specification. Six hours = 18,000 MWh ÷ 3,000 MW.
- Forest Advisory Committee minutes, May 2026, PDF pp. 97–99: Stage-I history and replacement compensatory-afforestation land.
- MPCB-hosted project executive summary, 2025: developer’s four layout alternatives (p. 13) and environmental assessment summary.
- Denver Water, tree-removal-plan amendment (2024), with the 2021 plan: amendment p. 2: initial clearing in 2022; original plan pp. 2–3: 415 forest acres and about 200,000 trees; amendment pp. 2–3: removal threshold.
- Dominion Energy, Bath County Pumped Storage Station: 3,003 MW net generation capacity and 1985 start. No original tree-clearing count established.
- U.S. Department of Energy, pumped-storage hydropower: operating principle.