.png)
A €3.6 billion German reactor spent 19 years under construction and was never switched on. It is now a theme park. Between 1944 and 2026, nine countries put over $100 billion into fast breeder reactors — machines designed to produce more fuel than they consume. France ran the world's only commercial-scale unit at under 7% capacity. Japan spent $8.5 billion on a reactor that generated power for less than a year. Only Russia operates one today.
Prof. Suneet Singh of IIT Bombay explains why the technology broke every country that attempted it. Fast breeders need liquid sodium, which burns in air and explodes in water. They need a breeding ratio above 1.0 sustained at commercial scale. And they need a reprocessing loop to close the fuel cycle. Every country that failed solved at most one. Russia had no better answer to any of them individually — it just treated the reactor as a system rather than a machine.
India holds 1–2% of the world's uranium and more than 25% of its thorium. Homi Bhabha designed a three-stage programme in the 1950s around exactly that asymmetry. Of seven available nuclear pathways, India has pursued every one that reduces uranium dependency and deprioritised every one that deepens it. Coal's costs are no longer absorbable. Solar fails for four monsoon months in the states holding the most capacity. The path to sovereign energy runs through thorium.
After Pokhran in 1974, the Nuclear Suppliers Group cut India off entirely. So Kalpakkam ran a 13.5 MW test reactor from 1985 and trained three generations of scientists who understood sodium as a material with a personality. Two hundred Indian industries built PFBR components because there was no one else to call. The economics still don't work — electricity will cost 80% more than from India's heavy water reactors, and Stage 3 doesn't exist anywhere on earth. Commercial penetration is 20 to 30 years out. That is the nature of a 70-year programme.
Unicorn India Ventures is an early-stage deeptech fund investing in semiconductors, space, defence, robotics, health, and agritech. We back companies where the hard part is the technology, and where the payoff sits on a longer clock than most capital is willing to hold. This series examines what it actually takes to build things that don't work for forty years and then do.
.svg){kind=small}
