Before Sodium Fast Reactors (SFRs) became LinkedIn and venture-capital darlings — before the visible euphoria that sweeps the room when someone says the magic words “Sodium Fast Reactor” — France actually built one. It was called RAPSODIE. RAPSODIE went critical in 1967 at Cadarache. An SFR operating at ~40 MW thermal, it never produced electricity; that wasn’t the mission. It was built to generate data on fast-spectrum physics, sodium systems, and materials behavior. And it did. It ran for 16 years and produced something enthusiasm alone cannot: Operating experience. Lots of it. Here’s what RAPSODIE actually taught us: 1.  Sodium works. It is not magic. Liquid sodium is an excellent heat transfer medium. High thermal conductivity. Low operating pressure. Elegant thermodynamics. It moves heat like a 100-car freight train with no speed limit. It is also chemically aggressive. As in...VERY chemically aggressive. Sodium reacts violently with air and water. Leak detection must be immediate. Fire protection must be engineered into the plant DNA from day one. Maintenance becomes controlled chemistry management under tight procedural discipline. Sodium isn’t just a coolant; it’s a hot live wire running through your primary system.

RAPSODIE proved sodium cooling is feasible. It also proved it demands industrial precision at the very highest level. Feasible does not mean simple. 2. Fast neutrons are hard on materials.

Fast-spectrum physics is beautiful. Fast-spectrum materials performance is unforgiving.

Fast neutrons don’t gently age steel — they sandblast it at the atomic level.

Swelling. Embrittlement.

Irradiation damage under high neutron flux and elevated temperature. RAPSODIE wasn’t just validating reactor theory. It was stress-testing structural materials over time. The data was invaluable — and it showed that durability in fast systems is a long-term engineering commitment, not a solved problem. Physics success does not equal lifetime confidence. 3. Experimental success does not equal commercial success. RAPSODIE informed Phénix and later Superphénix. But scaling up introduced complexity, cost, and operational realities that enthusiasm alone cannot erase. Prototypes prove concepts. They do NOT prove:- economics. - fleet reliability. - competitiveness in real energy markets. If we’re going to advocate SFRs in 2026, that’s fine. But let’s do it honestly. The historical record does not say “sodium is perfect and was unfairly abandoned."

It says:"- It works. -It’s demanding. -It stresses materials. It has a limited commercial operating history. "

Engineering maturity is earned through operating scars — not applause or clever slide decks.

The lesson isn’t that sodium is impossible.

The lesson is that sodium is NOT magic