One glacier collapse reached more than eleven Nepali hydropower sites

The missing-to-death ratios on each side of the border no longer describe the same event: Tibet reports 546 missing against 16 deaths, a ratio near 34 to 1, while Nepal reports 2,498 missing against 734 deaths, a ratio near 3.4 to 1 ^[20][1][14]. A tenfold gap in that ratio is either a difference in how a valley gives up bodies or a difference in how a register of the missing gets compiled, and until it resolves, the 750 headline is the least settled number in the file ^[19][14].

The number that matters for capital is the other one. At least 933 workers are missing from more than 11 hydropower sites, which averages under 85 a site and amounts to about 37 per cent of everyone Nepal currently counts as missing ^[3][16][17]; roughly 100 of the 933, about one in nine, are thought to still be alive underground ^[3][18].

A run-of-river asset earns from gradient, and gradient here means a steep valley with a tunnel cut into the side of it, which is where the workers were when the ice, rock, mud and debris came down the rivers ^[5][3]. More than eleven sites were hit by a single glacier collapse on a single Wednesday ^[4][3]. For a lender, that is eleven exposures tied to one hazard, a book that looks diversified on a schedule but concentrates risk within a single basin.

This is probably wrong, but I would expect the repricing to arrive as tunnel-occupancy rules and construction-phase cover rather than as abandoned sites, because the head is where the gradient is and no sponsor can move a river. The cheapest fix has in fact already been agreed at nearly no capex: China and Nepal said on the 29th they would expand real-time hydrological sharing ^[12], and Wang Yi said China had passed footage, satellite and hydrological data, and warnings about upstream barrier lakes ^[13]. That is where the hazard signal was living: upstream, across a border, on another ministry's desk.

What would prove the siting read wrong: if most of the 933 turn out to have been in surface camps and spoil yards rather than inside the tunnels ^[3], this becomes a question of where you house people in monsoon and not a question of where you put concrete. And if Nepal's toll settles toward the 675 the South China Morning Post carried rather than the 734 in the count relayed by en.sedaily.com from Reuters and the Associated Press, a gap of 59, roughly 8 per cent, the human total moves while the eleven sites stay exactly where they are ^[21][20][22].

The allocation choice is visible in the field, meanwhile, because rescue capacity is concentrated on Trishuli 3A and 3B ^[8], where the army drilled into the upper section of a buried tunnel in Bagmati Province before starting excavation ^[6], while the Bhote Koshi rises ^[2]. Nine or more sites are waiting their turn.