Nepal’s Floods Are a Warning Against a Mega Dam on the Siang
The disaster that tore through Nepal and Tibet last week was not a routine monsoon flood. On the morning of 26 August, a glacial collapse on the Chinese side of the border sent ice, rock, mud and debris into the Bhote Koshi–Trishuli river system. Within minutes, valleys were buried, bridges vanished, and hydropower tunnels became death traps. By the end of August, authorities were reporting around 800 confirmed deaths across Nepal and Tibet and more than 3,000 people still missing. Among them were hundreds of workers employed at hydropower plants along the same rivers that were supposed to generate “clean” power. The Red Cross estimated that more than 90,000 people had been affected.
That is the landscape against which India should now judge its plan to build a mega dam on the Siang River in Arunachal Pradesh.
What the mountain actually did
Officials first wondered whether a small earthquake had set off the flood. Subsequent analysis pointed the other way. A mass of ice and rock failed high in the Himalaya, slammed into a mountain river near the Rasuwagadhi–Gyirong corridor, and produced a debris flood. Temporary lakes formed behind the rubble and later overflowed. There was no preceding cloudburst of the kind that usually explains Himalayan flash floods. The trigger was geological: ice and slope failing in a steep, seismically active landscape.
Rescue work has been grim and slow. Nearly 20,000 security personnel have been deployed in Nepal. Helicopters have been grounded by weather. Bodies have been recovered far downstream, some in a condition that makes identification almost impossible. Mortuaries filled up; mass burials of unidentified victims began. The most haunting image of the week has been the search inside hydropower tunnels. Hundreds of workers remain unaccounted for at sites such as Chilime and the Trishuli projects. India sent a specialised tunnel-rescue team. An air pipe was inserted into at least one blocked tunnel in the hope that someone was still alive. The same infrastructure that was sold as development became a maze from which people could not escape.
This is not an isolated freak event. In October 2023, a glacial lake outburst at South Lhonak in Sikkim destroyed the 1,200 MW Teesta-III dam at Chungthang before its gates could be fully opened. The flood then raced hundreds of kilometres down the Teesta, taking bridges, settlements and farmland with it. In February 2021, a glacial collapse in Chamoli, Uttarakhand, wiped out the Rishiganga project and trapped workers in a tunnel. The Himalaya keeps delivering the same lesson: ice, rock and water do not respect engineering drawings.
The dam India wants to build
The Siang is the name the Brahmaputra takes as it enters India from Tibet, where it is the Yarlung Tsangpo. China is building a far larger project at Medog, near the Great Bend, with a planned capacity of around 60,000 MW. New Delhi and Itanagar have framed the Siang Upper Multipurpose Project as the answer: a storage dam 250–300 metres high, a reservoir of about 9 to 9.2 billion cubic metres, and 11,000–11,200 MW of generation, producing an estimated 47 billion units a year. Officials say it will create “prior use” rights, generate power, and absorb sudden releases from upstream. Cost estimates run from about ₹1.13 lakh crore to ₹1.5 lakh crore. Proposed sites have included locations near Geku, Ugeng, Dite Dime and Parong.
The strategic worry is not invented. An upstream neighbour that can store and release water at will is a genuine risk for Arunachal Pradesh and Assam. China’s project sits in a seismic zone, and New Delhi has received little usable hydrological transparency. Arunachal Chief Minister Pema Khandu has repeatedly said the Siang dam is not only about electricity but about protecting communities from unregulated or sudden discharges across the border. Some villages have signed memorandums of understanding for pre-feasibility work. The state government says it will not rush the study.
None of that makes a mega dam on the Siang a safe reply to Nepal’s warning.
Why the Nepal disaster cuts against the project
First, the geology is the same. The Siang valley is a high-seismic, high-slope, glacier-fed basin. A large reservoir does not cancel glacial collapse, landslide dams or cloudbursts. It can amplify them. If a debris pulse or glacial lake outburst hits a full reservoir, operators face an ugly choice: overtop the dam or release a wall of water into Assam. Sikkim already showed what happens when a Himalayan dam meets a sudden pulse it was not designed, or not ready, to pass.
Second, hydropower and flood control pull in opposite directions. To generate power, a reservoir is kept high. To absorb a flood, it must be kept low. A single structure cannot do both jobs well. A dam built mainly for energy will behave like many Himalayan projects before it: when the unexpected pulse arrives, it becomes part of the flood, not a shield against it. Downstream Assam has every reason to fear that a “safety” dam in Upper Siang could, in a crisis, dump water faster than the plains can bear.
Third, the people who would live with the dam have said no for years. Adi communities call the river Aane Siang — Mother Siang. The Siang Indigenous Farmers’ Forum and the Adi Banne Kebang have opposed surveys as well as construction. Protest estimates put at least 27 villages under water and many more disrupted; other figures run far higher. Exact submergence is still unknown because full, trusted surveys have not been completed. That uncertainty is itself a warning. A project whose physical footprint is still contested should not be sold as a finished national security solution. A trust deficit this deep is not a public-relations problem. It is a safety problem. Projects built against the people who know the slopes and the river are the ones that cut corners on siting, silt and emergency plans.
Fourth, Nepal’s tunnels should end the romance of underground Himalayan hydro. Last week’s missing workers were not an accident of poor luck. They were the predictable result of putting large workforces inside mountains beside unstable rivers. The Siang project would require similar excavation in similar rock. That is not an abstract risk. It is last week’s news.
What a serious response would look like
The Nepal floods do not prove that China’s Medog dam caused this disaster. That project is still under construction. They do prove something more basic. In this mountain system, ice can fail, debris can block a river, a temporary lake can burst, and hydropower infrastructure can become a death trap in a single morning.
Building an even larger storage dam on the Siang does not cancel that physics. It adds a new failure mode — a man-made lake in a fragile gorge — on top of the natural ones. It also locks India into the same model that just failed in Nepal and, three years earlier, in Sikkim: treat the Himalaya as a power plant first and a living, unstable landscape second.
A serious response to China’s upstream plans would look different. It would mean real-time data sharing that actually works, not a mechanism that goes quiet when relations freeze. It would mean early-warning systems that reach villages in hours, not after the flood has already passed. It would mean smaller interventions where the geology allows them, and floodplain management in Assam that does not pretend a single wall of concrete can hold the Brahmaputra. It would mean listening to the Adi when they say the river is not a vacant site for a national project.
Nepal has just shown what happens when the mountain moves faster than the plan. India should not spend more than a lakh crore rupees to repeat the experiment on the Siang.
About The Author
