Nanda Devi from the south-west: the summit, the south-west face and the ridges at its foot, rendered from satellite heights and a satellite picture

Nanda Devi7,816 m

No. 10 · Nanda Devi, Uttarakhand · 5 October 2026

Where is the generator lost on Nanda Devi in 1965?

A plutonium-powered generator was tied to a ledge near 7,200 m in October 1965. By spring the ledge was gone. Twelve pictures on where sixty-one years of gravity and ice have most likely put it.

Rendered view, not a photograph · Esri World Imagery (Esri, Vantor) · captured 15 September 2018, over terrain from the High Mountain Asia 8 m DEMs and Copernicus DEM GLO-30 · from the south-west · heights not exaggerated

02 / 12 · the load

What was carried

A SNAP-19C generator: seven capsules holding 1,900 g of plutonium alloy, 1,386 g of it plutonium-238, built to give 40 W of electricity for two years.

Prime Minister Morarji Desai told the Lok Sabha on 17 April 1978 that the decision was taken “at the highest level” of both governments, India’s and the United States’.

Drawn to scale · casing as stated to Parliament, capsules as measured in the 1979 report

Report of the Committee of Scientists, 1979 · Lok Sabha, 17 April 1978

03 / 12 · the camp

The ridge

The route was the south ridge, first climbed in 1936. The fourth camp sat in “a small rocky depression” used by every party before, about 600 m below the summit.

The 1965 party’s own figures put it at 7,224 to 7,239 m; other parties give 7,010 to 7,163 m for the same site. On the crest that is a stretch of 330 m, marked white.

Rendered view, not a photograph · Esri World Imagery (Esri, Vantor) · captured 15 September 2018, over terrain from the High Mountain Asia 8 m DEMs and Copernicus DEM GLO-30 · from the west-south-west · heights not exaggerated

Esri World Imagery (Esri, Vantor) · High Mountain Asia 8 m DEMs (Shean 2017, NSIDC) · Copernicus DEM GLO-30 · Tilman 1937 · Report of the Committee of Scientists, 1979 · Kohli and Conboy, Spies in the Himalayas (2003)

Nanda Devi from the south-west with the south ridge traced to the summit and the stretch of crest where the fourth camp stood

Summit7,816 mFourth camp7,010 to 7,239 mSouth ridge

04 / 12 · what was found

Spring 1966: the shelf is gone

“The rock shelf where he had left the boxes was completely gone, as if a massive knife had surgically severed that one corner of the slope.”Kohli and Conboy, Spies in the Himalayas (2003)

The official account says the same: “the entire rock where these things were kept seemed to have broken”. A climber and four Sherpas reached the camp on 1 June 1966. A collapsed tent, a gas canister, a radio and lengths of wire were there. The generator and the rock it was tied to were not.

Drawn from the accounts; not to scale, and the shape of the ledge is not known · inset: rendered view, not a photograph · Esri World Imagery (Esri, Vantor) · captured 15 September 2018, over terrain from the High Mountain Asia 8 m DEMs and Copernicus DEM GLO-30

Kohli and Conboy, Spies in the Himalayas (2003) · Report of the Committee of Scientists, 1979, Annexure 3 · inset: Esri World Imagery (Esri, Vantor) · High Mountain Asia 8 m DEMs (Shean 2017, NSIDC) · Copernicus DEM GLO-30

Nanda Devi from the west-south-west, with a ring on the south ridge where the fourth camp stood — **Where this is**The fourth camp, on the south ridge, about 600 m below the summit.

05 / 12 · gravity

Two ways down, and a third

A camp on a crest can lose a load down either side. West of the crest the ground sends a fall south-south-west onto the glacier, up to a crest height of 7,322 m; higher than that it goes west. East of the crest everything runs to the glacier’s head, out of sight behind the ridge.

south-south-west40 %

west5 %

south-east, behind the ridge55 %

by the shape of the ground alone

Rendered view, not a photograph · Esri World Imagery (Esri, Vantor) · captured 15 September 2018, over terrain from the High Mountain Asia 8 m DEMs and Copernicus DEM GLO-30 · from the south-west · heights not exaggerated · paths: steepest descent with an avalanche runout rule; dashed where ground stands in front

Esri World Imagery (Esri, Vantor) · High Mountain Asia 8 m DEMs (Shean 2017, NSIDC) · Copernicus DEM GLO-30

Fall paths from the fourth camp drawn on the south-west face: most run south-south-west to the glacier, some west to the valley floor

Fourth camponto the glacierto the valley floor

06 / 12 · evidence

The test few remember

On 8 June 1966 the search party dropped a part-used gas cylinder from the camp. It was found “in the glacier just on the way between the sanctuary and base camp”. Only the south-south-west fall ends there.

south-south-west89 %

west4 %

south-east7 %

with the cylinder weighed in · a judgement, not a measurement: other reasonable weights give 75 to 92 %

Rendered view, not a photograph · Esri World Imagery (Esri, Vantor) · captured 15 September 2018, over terrain from the High Mountain Asia 8 m DEMs and Copernicus DEM GLO-30 · from the south-west · heights not exaggerated

Report of the Committee of Scientists, 1979, Annexure 3, p. 18 · Esri World Imagery (Esri, Vantor) · High Mountain Asia 8 m DEMs (Shean 2017, NSIDC) · Copernicus DEM GLO-30

The south-south-west fall path from the camp to the glacier, the way the dropped cylinder went

Dropped here8 June 1966Found on the glacier

07 / 12 · one lead, closed

The box that wasn’t

In 1993 an Army expedition carried nearly a ton of old rubbish off the mountain. In it was the olive-drab box that had brought the capsules up in 1965.

It “was whisked away to check for radioactivity. Tests proved negative.” It had been left empty at base camp. It says nothing about where the generator went, and it is the only test we find on record since 1979.

Dates from the parliamentary record and the 1979 report; 1993 from the expedition leader’s account

Kohli and Conboy, Spies in the Himalayas (2003) · Report of the Committee of Scientists, 1979

1965October The generator is tied to a rock shelf beside the fourth camp as a blizzard drives the party down.
19661 June The shelf is gone. Searches on foot and by helicopter run until late 1968.
197817 April The Prime Minister tells the Lok Sabha.
197918 May The Committee of Scientists reports: still “undetected and unretrieved”.
1993An Army expedition brings out the box that carried the capsules. It tests negative.
2021The atomic research centre answers an information request: it “has no information”.
2026Sixty-one years on the mountain.

08 / 12 · sixty-one years

The ice barely moves

Satellites have timed this glacier since the 1980s: 18.6 m a year above the bend, 7.6 under the face, 3.2 on the lower tongue.

Anything that landed on the ice under the face has travelled about 203 m between 1965 and 2026. Across the runs, 22 to 539 m.

Rendered view, not a photograph · Esri World Imagery (Esri, Vantor) · captured 15 September 2018, over terrain from the High Mountain Asia 8 m DEMs and Copernicus DEM GLO-30 · from the south-west, across the glacier · heights not exaggerated · white band: where a fall from the camp reaches the ice · ice speed: 120 m grid, all years

ITS_LIVE v2.1 (NASA MEaSUREs) · Esri World Imagery (Esri, Vantor) · High Mountain Asia 8 m DEMs (Shean 2017, NSIDC) · Copernicus DEM GLO-30

The lower glacier seen across from the south-west, the face rising behind it, its centreline marked every kilometre and the short distance the ice has moved since 1965

3 km4 km5 km6 kmice stops moving1965 → 2026

09 / 12 · the answer, as far as it goes

Where that leaves it

Most likely in or under the lower glacier, 4.5 to 5.4 km down its centreline. About 7 runs in ten end here; the rest stop higher, on the face or in the gullies searched in 1966–68, or on the other sides.

The generator gave 800 W of heat in 1965 and about 490 W now: on ice it melts its way to the bed. The ice stops moving 1.3 to 2.2 km further on. At its present pace that is five centuries or more away.

Esri World Imagery (Esri, Vantor) · captured 15 September 2018 · 0.5 m source, shown at 1.25 m to the pixel · north up

Esri World Imagery (Esri, Vantor) · ITS_LIVE v2.1 (NASA MEaSUREs) · Report of the Committee of Scientists, 1979, Annexure 7

The lower glacier from directly above at half-metre grain, with the stretch where the generator most likely lies marked in red

3 km4 km5 km6 kmMost likely hereat the bed of the iceice stops movingthe fall comes down here

1 kmN

10 / 12 · downstream

What flows from here

The glacier’s water leaves the Sanctuary as the Rishi Ganga, joins the Dhauliganga just below Reni, and the Dhauliganga joins the Alaknanda at Vishnuprayag. At Devprayag the Alaknanda meets the Bhagirathi and becomes the Ganga.

By ice, anything in the tongue is centuries from its end. By water under the ice it could be one melt season, if a channel connects. Nobody has measured the drainage under a debris-covered Himalayan glacier.

glacier
Rishi Ganga
Dhauliganga
Alaknanda
Ganga

Rendered view, not a photograph · Esri World Imagery (Esri, Vantor) · captured 15 September 2018, over terrain from the High Mountain Asia 8 m DEMs and Copernicus DEM GLO-30 · from the south-east, looking down the gorge · heights not exaggerated · rivers as mapped in OpenStreetMap; dashed where a ridge stands in front

© OpenStreetMap contributors · Chamoli district, Government of Uttarakhand · Report of the Committee of Scientists, 1979, Annexure 7 · Esri World Imagery (Esri, Vantor) · High Mountain Asia 8 m DEMs (Shean 2017, NSIDC) · Copernicus DEM GLO-30 · Miles and others 2017

The Nanda Devi Sanctuary and the Rishi Ganga gorge from high in the south-east, the river traced from the glacier to its meeting with the Dhauliganga at Reni

Nanda Devithe glacierRishi GangaRenijoins the DhauligangaVishnuprayagjoins the Alaknanda

11 / 12 · monitoring

Who is measuring?

In 1979 the Committee of Scientists judged the hazard “to the public at large” to be “negligibly small in magnitude”, noted that the device “still remains undetected and unretrieved”, and asked for continuing monitoring of air, water, soil and river sediment. The government accepted.

Since then we find one test on record, the 1993 box, and no published series. In 2021 the Bhabha Atomic Research Centre answered an information request: “This center has no information on the above.”

Decay arithmetic only: it says how much is left, not where it is or what dose anyone receives

Report of the Committee of Scientists, 1979, para 4.6 and memorandum of action · New Indian Express, 17 Feb 2021

12 / 12 · limits, method, sources

What we cannot say

Which side of the crest the loads sat on. No account says so outright.
Whether it broke up. The 1979 report lists six possible fates; none is ruled out here.
Anything about dose. No satellite sees the object or its radiation. This is about place.
The shares are not measured odds. They lay out what the ground and the record allow.

How it was made. The camp was placed on the ridge from the expedition accounts and the 1979 report. Falls were run downhill from it on 30 m and 8 m elevation models and stopped by a published avalanche runout rule. What reached the ice was moved for 61 years at the speeds satellites measure. The 1966 drop test was then weighed in.

Esri World Imagery (Esri, Vantor) · captured 15 September 2018 · 0.5 m source, shown at 1.5 m to the pixel · north up · orange: the fall to the south-south-west · ring: where a fall to the west would come to rest, by the runout rule

Maplore Gazette · counted and drawn by Maplore

Everything on one map: the camp on its ridge, the fall to the glacier, the ring where a fall to the west would end, and the stretch of glacier where the generator most likely lies

Nanda DeviFourth campmost likely herea west fall would end hereon the valley floorice stops moving

1 kmN