top of page

What Nepal's Floods Reveal About Grid Concentration Risk

Aug 31
3 min read
Tamakoshi River was diverted, tunnels were cut through unstable rock, and a 456 MW hydroelectric dam was built in Nepal. (en.clickpetroleoegas.com.br)
Tamakoshi River was diverted, tunnels were cut through unstable rock, and a 456 MW hydroelectric dam was built in Nepal. (en.clickpetroleoegas.com.br)

On 26 August 2026, a glacier collapse in Nepal's Langtang region dammed a mountain river with ice and rock. When the temporary dam failed, the surge travelled nearly 100 kilometres downstream, destroying homes, roads, a border highway, and several hydropower stations along the way. Nepal's Electricity Authority has confirmed the floods knocked 431 megawatts of operating hydropower capacity off the national grid, with a further 470 megawatts of capacity under construction also damaged — together, close to a tenth of the country's generating capacity, lost within hours.


It was not a rainstorm. Scientists examining the event have pointed to a slope failure connected to warming and permafrost degradation at altitude, a hazard category that conventional grid and flood planning were not built around.


Concentration, not just capacity


The immediate story is about lost megawatts. The more useful engineering story is about geography. In a hydropower-dependent grid, generation stations, transmission corridors and substations tend to sit in the same river valleys, because that is where the resource is. Under normal conditions, that is an efficient way to build a power system. Under a single geological event, generation and transmission can fail together because they were never meaningfully separated to begin with.


This pattern is not confined to Nepal or to hydropower. Any grid architecture that routes generation, distribution and backup fuel supply along the same physical corridors carries the same structural exposure — to floods, wildfires, ice storms, or any other event that follows a single geography. Nepal's own 2024 flood season damaged eleven hydropower stations and dozens of roads and bridges in one event; the pattern in 2026 is a repetition, not an anomaly.


Why backup power alone doesn't solve it


The standard response to an event like this is to call for more backup capacity at critical sites — hospitals, water treatment plants, emergency communications hubs. That is the right instinct, but it depends on the backup itself not sharing the same failure points as the system it is meant to replace. A diesel generator needs a fuel truck, and the fuel truck needs a road that may no longer exist. A battery system holds only what it was last charged with, and stops helping the moment the outage outlasts the charge.


Critical infrastructure needs energy that can be positioned locally in advance, held without loss over time, and released only where and when it is needed — independent of both the grid connection and the fuel logistics chain a disaster is most likely to break first.


A different way to hold energy in reserve


This is the problem iQ-LOOP's platform addresses directly. Energy is stored in aluminium — a stable, transportable solid that holds its energy content indefinitely with no standing loss — and released as heat, power and hydrogen only at the point of use, through our CREATE process. The aluminium moves as inventory between sites; the hydrogen itself is never transported or stored separately, because it is generated on demand, on site, at the moment it is needed.


For a hospital or emergency shelter, that means the energy reserve on site does not depend on a live grid connection, a full fuel tank, or a road that survived the disaster. It depends on inventory that was already there.


No energy architecture prevents a glacier collapse. What changes is what happens in the hours and weeks after: whether critical facilities downstream keep functioning while the grid is rebuilt, or whether they are waiting on the same fragile chain that just failed.


The planning question this leaves us with


Nepal's disaster agency and international researchers are still investigating the full technical cause of the Langtang collapse. But the infrastructure question it raises stands regardless of the final answer: how much of what gets called grid resilience today has actually only been tested under the conditions it was designed for.


Grids built for the climate of the last century are increasingly being tested by a climate that no longer follows those assumptions. Decentralised, on-site, non-grid-dependent energy belongs in core infrastructure planning from the outset — not as an emergency add-on reached for once the next event has already made the case.


*iQ-LOOP is a circular aluminium energy platform delivering clean heat, power and hydrogen on demand at the point of use — without the grid dependency, fuel logistics, or hydrogen storage and transport that conventional systems require. Learn more at www.iq-loop.co.uk.* 


 
 
 

Comments


bottom of page