The Siang Upper Multipurpose Project is still at the pre-feasibility stage, with no dam site finalised and not a single borehole drilled. Yet it has already become one of the most bitter arguments in the Northeast. It pits a government worried about a giant Chinese dam upstream against the Adi people, one of the largest indigenous groups in Arunachal Pradesh who fear losing their villages, farms and a river they regard as their mother.
This old soldier spent enough years on the China front—including commanding a brigade in Arunachal’s Tawang—to feel the pull of both arguments. Therefore, this article is an attempt to do what the shouting on each side has made hard: taking both cases seriously and saying plainly where each holds.
Let’s start with the government’s worry. China approved its 60,000-MW Medog dam on the Yarlung Tsangpo in December 2024 and began work on July 19, 2025, roughly 50 km from where the river enters India, without prior consultation with the downstream countries.
India has no water-sharing treaty with China, and the flood-season data agreement that lapsed in June 2023 is still being renegotiated. Nobody in Pasighat, headquarters of East Siang district, needs to be told what a surprise from upstream looks like, because in June 2000 a dam burst in Tibet resulted in the Siang river rising nearly 14 metres above the danger level.
Arunachal Pradesh Chief Minister Pema Khandu calls Medog an existential threat, and the Union government has pitched the Siang project as a buffer that can regulate floods. Frankly though, the “water bomb” language needs trimming, since roughly 65-70 percent of the Brahmaputra’s annual discharge is generated inside India; China describes Medog as a run-of-the-river scheme.
The sharper risks are operational—an unannounced release in monsoon season or a squeeze in winter, when the Central Water Commission is cited as putting China’s share of the Siang’s flow at about 84 percent. Those are good reasons to want data and warning time, and the government is right to want both.
Now, let’s do the maths. At 11,200 MW, the Siang project is less than a fifth of Medog, so it cannot counter the Chinese dam in any like-for-like sense. Its honest case is storage and signalling. About 9.2 billion cubic metres of live storage could, in principle, absorb a sudden surge and sustain lean-season flows, though that depends on design and operating rules yet to be settled, and the project tells Beijing that India means to build on a river flowing through territory China claims.
Now for the other side that deserves the same respect. For the Adi, the river is Ane Siang—Mother Siang—and their opposition is no passing mood. The community’s council rejected the first proposal in 1979 and the Bogum Bokang Kebang, the community’s supreme inter-village decision-making assembly, met at Geku on August 20, 2025—for the first time in about 25 years—to reject this one.
Their fears are not invented. A study this March by Tadu Rimi of Shiv Nadar University estimated that 27 villages could be submerged and 43 more affected, and a wall of 280-300 metres is being contemplated in a part of the Himalayan arc that the revised seismic map of 2025 puts in a new highest-hazard zone.
The hardest question comes from a former president of the Adi Bane Kebang United, Thomas Borang, who asks whether a large dam of India’s own would not do double the damage if an earthquake, a missile or ageing concrete brought it down. The government’s answer—that a regulated reservoir turns an uncontrolled hazard into a managed one—is a legitimate engineering position. But it hasn’t yet been shown through a site-specific seismic hazard assessment and a dam-break study.
In fairness, not every argument on this side holds in its strongest form either. The claim that Medog poses no real danger sits uneasily with reports of about 5.5 billion cubic metres of storage there and with Chinese geologists’ warnings about an active fault under the site.
The Adi are not of one mind either, since some villages and households have signed memorandums in support of the pre-feasibility study. That is a reason to ask people properly, village by village, rather than assume a verdict in either direction.
To be fair to the government, it has not been silent. Khandu has repeatedly said that no decision has been taken and that nothing will go ahead without consent. In June, the Union government and the state announced a `350-crore programme for 26 villages likely to be affected.
Then came October 23, when a bandh called for the release of four detained protesters turned violent. The National Hydroelectric Power Corporation office at Pasighat was set on fire, vehicles were burned and the homes of two legislators damaged.
No government can shrug off arson. Equally, this old soldier knows from experience that the size of a security presence sends its own message. Thirty companies of central armed police forces, anti-terror law UAPA invoked against protesters and an internet shutdown that has run for weeks add up to a heavy response to a dispute that began over survey work. Villagers have started calling it an ‘occupation’.
To its credit, the state has begun to ease off. On November 7, the four detained protesters were released from Jully central jail as a goodwill gesture, suspended village headmen are being reinstated, and Khandu has invited the Siang Indigenous Farmers’ Forum to name a delegation and a date for talks. One activist called it a partial victory, since dozens of others remain in custody and movement leaders want the cases withdrawn and the forces pulled back.
The project was argued as a matter of national security before the public had seen a single site-specific study, and that order of events has cost trust, while a flat refusal to allow even a study leaves no room for the evidence that could settle the safety question one way or the other.
What could be a sensible way forward? Here are four steps, none of which asks anyone to lose face.
First, revive hydrological data sharing with China and put gauges and early-warning stations upstream whatever happens to the dam, because a flood wave takes about 12 hours to travel from Pasighat to Dibrugarh—enough time to save lives if somebody sees it coming. Second, publish the seismic, dam-break and sediment studies with independent review before any decision is taken. Third, treat the pre-feasibility work as a study and not a verdict, with no drilling where the village councils haven’t agreed and a gradual drawdown of forces as talks progress. Fourth, accept what the studies say, whichever way they point, and tell the people of the valley first.
There is a larger point as well. The marchers who set out for Itanagar last week sang the national anthem before they left, which says a good deal about how they see themselves; Chinese commentators are already noting their unhappiness.
India can afford to take this slowly because, even on official timelines, the first electricity output is years away. What it cannot afford is a dam or a debate that leaves its own border people feeling like bystanders on their own river.
Maj Gen Sudhakar Jee, VSM (Retired) | Commentator on geostrategy who commanded a brigade in Tawang and an infantry division in Ladakh
(Views are personal)