Nine Days Buried, Then Voices Answered: Nepal Tunnel Rescue Defies 1,380 Dead

What happened on Friday, September 4, 2026, nine days after the August 26, 2026 flash flood and ice-rock-mud slide that swept the Trishuli/Bhotekoshi valley on the Nepal-Tibet border, was centered on the buried Trishuli 3A Hydropower Project tunnel in Rasuwa district⁠2⁠3⁠4⁠8. Nepal Army rescuers working the collapsed portal shouted into the void and, for the first time since the August 26 flood, a voice answered back⁠10. The Times of India gives the exchange verbatim, and it is worth reading plain: rescuers called out, and heard back the Nepali affirmation that they were heard⁠10. By 9:20 a.m., two workers were out⁠10. France's AFP news agency called it a "miracle" on the tenth day of rescue efforts, and framed it as a rare "ray of hope"⁠4.

The Friday rescue has a name, or rather two names — and the record needs care here because sources disagree on who was saved when. The BBC and AFP name the September 4 pair as Sanjaya Shah, also rendered as Sanjay Sa, and Kabir Maharjan⁠2⁠4. Al Jazeera and PBS then report a different pair rescued on Saturday, September 5: a Chinese national man and a Nepali woman, in separate operations⁠5⁠7. CNN likewise reports two further survivors rescued on Saturday, naming Chandika Shrestha, a woman rescued from Betrawati⁠9. In other words, the first miracle was followed by a second: at least four people pulled alive over two days, from different wreckage, not one repeated telling of the same two faces⁠5⁠7⁠9.

The confusion is understandable. For ten days the valley has produced almost nothing but missing lists, and any live rescue instantly becomes the story. But the distinction matters for families waiting at tunnel mouths. The Friday pair came from inside the Trishuli 3A tunnel in Rasuwa after nine days underground⁠2⁠3⁠4⁠8. The Saturday pair came a day later, ten days after the flood, from flood wreckage and a separate site at Betrawati, according to Al Jazeera, PBS and CNN⁠5⁠7⁠9. None of the provided sources merges them into a single event, and none explains how either group survived that long without food, light or clean air⁠5⁠7⁠9.

If the rescues are the sharpest fact, the death toll is the blurriest — and sources disagree by hundreds, a sign of how fast the count is moving and how incomplete the reconciliation remains. The BBC on September 4 put the toll at more than 1,200 killed, a figure matched that same day by Reuters via BusinessWorld, which also reported tens of thousands displaced⁠2⁠8. Al Jazeera on September 5 said the toll had passed 1,331, with missing on both sides of the border⁠5. The Straits Times on September 7 reported at least 1,380 killed and more than 5,500 missing in Nepal and Tibet combined, as Nepal observed a national day of mourning on September 7⁠11. No single number can be declared final, and to pick one would mislead⁠2⁠5⁠11.

What is consistent across those competing counts is scale: this is not a localized tunnel accident but a corridor-wide destruction of Nepal's hydropower belt. The Kathmandu Post reported on September 5 that 11 hydropower projects and a solar plant were damaged, with 942 people including project workers still missing at that point⁠6. Rescue missions continued at flooded hydropower sites, with teams searching for hundreds missing⁠6⁠7. Officials said 121 hydropower workers could still be trapped in tunnels as of September 7⁠11. That official figure is lower than claims circulating of more than 150 trapped in half a dozen tunnels, a higher number that no provided source confirms⁠11. The gap is not academic. It is the difference between how many air pockets crews think they are digging toward, and how many families are still being told to wait⁠6⁠11.

The Trishuli 3A rescue also reframes what survival required. Nine days without rescue in a tunnel sealed by flood debris means no light, limited air, cold water and no way to know if anyone knows you are there⁠2⁠4⁠10. The first voice heard from inside since August 26 — that answered call — is why AFP reached for words like "miracle" and "ray of hope" rather than routine rescue language⁠4⁠10. There are no other verbatim official quotes in the provided sources, no detailed medical readouts, no explanation yet of how Shah and Maharjan, or the Saturday pair, endured⁠5⁠7⁠9. What exists is the fact of extraction, and the continuation of missions around them⁠6⁠7.

For now, the honest position is narrow. Two were saved Friday, two more Saturday, and at least 121 may still be inside⁠2⁠5⁠11. More than 1,200 are confirmed dead by every count, with the latest counts pushing toward 1,380 and higher as missing lists are reconciled⁠2⁠8⁠11. Eleven hydropower projects are damaged, nearly a thousand workers and others were still listed as missing on September 5, and rescue teams have not stopped⁠6⁠7. Much of what circulated in the first chaotic days — precise foreign-national counts, seismic readings, gauge timings, corridor distances — does not appear in the corroborating reports provided here and cannot be stated as established fact⁠6⁠11. What can be stated is what the rescuers shouted, what answered, and that for at least two mornings in a row, the answer kept coming⁠10.