BLOG POST

Lessons from the Nepal Dam Disaster: Fix, Don’t Abandon Hydropower

In response to August's tragic floods in Nepal which killed dam workers along the Trishuli River, International Rivers and 121 other organizations have written to the World Bank Group calling for an immediate moratorium on large and cascade hydropower across the World Bank Group, and diversification away from hydropower for countries like Nepal. The tragedy does demand a response, but not one that denies sustainable development opportunities to countries rich in hydro potential.

On August 26, a massive rock and glacier ice slide sent a torrent of debris down the rivers of central Nepal. At least 1,451 people died. Among the dead and missing were 121 people trapped in hydropower plants, including about 50 at Upper Trishuli-1. The dam project was close to completion, but most of the upper dam was destroyed.

Upper Trishuli-1 is a project with significant World Bank Group involvement. The IFC co-developed the dam from 2012, holding a 15 percent stake. It financedand arranged $453 million in debt. MIGA guaranteed the Korean sponsors against breach of contract by the government. The IDA Private Sector Window provided $103 million in subsidized finance from the World Bank's soft-lending arm through IFC and MIGA, and the World Bank had earlier provided considerable technical support and policy guidance.

I have previously raised concerns about Upper Trishuli around the fact it was selected non-competitively and developed from an unsolicited proposal, and that there were potential conflicts of interest between the World Bank Group's various roles. I have no reason to think World Bank conflict-of -interest issues contributed to the tragedy. But there were failures that should be fixed.

The Trishuli hazard analyses were weak. Not least, more than 60 percent of the river's catchment lies in Tibet, but the project's 2018 environmental assessment drew its cumulative impact boundary at the Chinese border. Again, the assessment did not consider a rock-ice avalanche even though an avalanche in 2015 had briefly blocked a river that joins the Trishuli just above the dam. And this despite the fact that World Bank–funded dam safety guidelines for Nepal specifically covered mass movement hazards, including debris flows and landslides.

More broadly, the project fell short of World Bank best practice on dam safety. For large new dams, the World Bank has required reviews by an independent panel of experts of the investigation, design, construction and start-up phases, plus dam safety plans including an emergency preparedness plan. Both a co-financier (the Asian Development Bank) and the government also required an independent review panel (the government made the requirement for Trishuli in 2018). But it appears Upper Trishuli's independent panel only made its first site visit and report in June 2023, about two years into construction, leaving little time to fix issues. Once functioning, the panel rapidly requested a climate resilience assessment covering disaster risk, but this was incomplete in 2025. Again, in 2018 the IFC reported that a "standard dam break study" had been done, but by May 2025 the panel said dam-break analysis was still incomplete and was a priority. It also suggested that an emergency plan and early warning for the underground works were "urgently required." (Regarding earlier analyses, it appears glacier-flood and landslide risk assessments for the multilateral banks simply relied on the developer's environmental impact assessment.)

In response to these problems, the World Bank Group should ensure future projects meet two requirements:

  • As suggested by International Rivers and co-signatories, the World Bank Group should not sign off on financing for large hydropower projects before full safety and environmental studies are completed. This should include fully independent reviews of hazards across whole river basins and modeling cumulative impact, starting at the design stage, with reports to financiers rather than developers. There is no need for this to slow project development timelines if it is built in from the start of projects.
  • There should be no IFC or MIGA investment in infrastructure that doesn't follow World Bank best practice. IDA Private Sector Window-backed projects like Upper Trishuli in particular should meet the standards IDA would want if its money went through a government. And all IFC- and MIGA-financed projects should not only meet World Bank guidance on competition and transparency, but also on infrastructure safety standards.

Large dams are not “maladaptation” as suggested by International Rivers (which has long opposed World Bank Group finance for scaled hydropower), nor should the Bank Group stop financing them. Not least, hydropower has helped take Nepal's electrification rate from about 10 percent to nearly 100 percent, and made it a net power exporter. Reliable energy is an underpinning of economic growth, and hydropower is a comparatively safe source. The response to the tragic events at Upper Trishuli should be better analysis and processes, not denying Nepal and other countries the power they need for development. But to ensure it can continue to support such investment, the World Bank Group must do better at following its own advice on what works to deliver safe, timely, and affordable infrastructure.

DISCLAIMER & PERMISSIONS

CGD's publications reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions. You may use and disseminate CGD's publications under these conditions.


Thumbnail image by: Asian Development Bank/Flickr