Sealed Reactor Leaks Beside Live Warheads

Underwater shipwreck with encrusted steering wheel — Photo: AlDigital On / Shutterstock

The most dangerous thing about the Komsomolets is not that it leaks — scientists have known that for three decades — but that the leak sits inside a sealed room nobody is allowed to open, a mile beneath the Norwegian Sea, next to two nuclear warheads whose condition nobody can verify.

Key Points

A 2019 Norwegian-Russian expedition measured cesium-137 in water drawn directly from the wreck’s ventilation pipe at levels roughly 800,000 times normal Norwegian Sea background.
Norwegian authorities and the researchers who found the contamination both say it poses no demonstrated threat to people, fish, or the wider marine ecosystem, because the deep water dilutes it rapidly.
The reactor leak is confirmed and recurring, documented in the 1990s, 2007, and 2019; separate concerns about plutonium escaping the torpedo warheads remain unconfirmed by any published sampling.
Soviet and Russian designers patched the forward hull with sealant and titanium plugs in the mid-1990s on an estimated 20-to-30-year design life — a window that has already closed or is closing now.
Russia controls the wreck as a war grave and a military asset, which has limited independent inspection and halted joint monitoring with Norway since 2022.

What Actually Happened at 1,680 Meters

K-278 Komsomolets was a one-of-a-kind vessel: a titanium-hulled Soviet attack submarine, NATO-designated Mike class, built to dive deeper and faster than anything in the American or Soviet fleets. In 1984 it set a record for an operational combat submarine, descending to roughly 3,350 feet. Five years later, on April 7, 1989, an electrical fire broke out in the seventh compartment, fed by a ruptured high-pressure air line. The crew scrammed the reactor and blew ballast to surface within minutes, but the fire compromised the stern seals. The boat sank stern-first into the Norwegian Sea, settling at a depth of roughly 1,680 meters — about 5,500 feet. Of the 69 men aboard, 42 died, most from hypothermia while awaiting rescue that Cold War secrecy helped delay.

The wreck came to rest with its reactor core and two nuclear-tipped torpedoes, each carrying several kilograms of weapons-grade plutonium, still inside the hull. The titanium pressure vessel survived depths that exceeded its rated crush depth, which is the main reason the submarine is still a coherent object on the seabed rather than a debris field — and also why decades of proposals to raise it have gone nowhere.

The Confirmed Leak: What the 2019 Sampling Actually Showed

In July 2019, a joint Norwegian-Russian expedition aboard the research vessel G.O. Sars used a remotely operated vehicle to draw water directly from a ventilation pipe on the wreck — the same duct where Russian researchers had already documented releases in the 1990s and again in 2007. Onboard analysis found cesium-137 activity ranging from below 8 Bq/L to as high as 857 Bq/L in that sample stream, against an ordinary Norwegian Sea background of roughly 0.001 Bq/L. That is the source of the now-widely repeated figure that radiation near the wreck measured roughly 800,000 times normal levels. Norway’s Institute of Marine Research concluded plainly that releases from the reactor were still occurring thirty years after the sinking.

Why “800,000 Times Normal” Is Not the Same as “Dangerous”

Here is where the story gets misread most often. The eye-catching ratio compares an extraordinarily low background concentration against a sample taken at the exact point of release, inside a confined duct. It is not a measurement of the surrounding Norwegian Sea, and the same expedition found that other samples from that pipe were far lower. Norway’s radiation safety authority stated directly that the releases pose no risk to people or the marine environment, and the Institute of Marine Research’s expedition leader said the elevated readings were confined to the duct itself, not detected in the surrounding water column. At nearly a mile down, ocean currents dilute whatever escapes quickly enough that no measurable buildup of radionuclides has been documented in nearby sediment or biota.

That is a genuine scientific consensus, not spin: a confirmed, recurring release at the source can coexist with negligible impact on the wider ecosystem, because dilution at depth does the work that distance does near the surface. The unresolved question is not whether cesium is escaping — it plainly is — but whether that rate could change if the hull’s internal pathways degrade further, and whether a parallel plutonium release from the warheads is also underway.

The Plutonium Question, and Why Russia’s Own Designer Downplayed It

The more alarming possibility, raised repeatedly by environmental groups since the early 1990s, is that seawater reaching the torpedo compartment could eventually corrode the two nuclear warheads and release weapons-grade plutonium — a far longer-lived and more biologically hazardous contaminant than reactor cesium. Igor Spassky, the designer at Russia’s Rubin Institute responsible for Komsomolets, has consistently rejected the catastrophic framing. In 1993 he told the BBC it was “not a catastrophic situation”, and in a 1994 U.S. Naval Institute interview he argued that near-freezing water at the site would slow warhead corrosion for decades and that radioactive particles would have to pass through an internal labyrinth before reaching the ventilation trunk. Subsequent sampling has, so far, tended to support that caution: a 2019 study found no plutonium from the torpedo warheads in surrounding water and judged the titanium patches installed in the mid-1990s still effective, and a separate Norwegian-Russian comparison likewise found no evidence of weapons-grade plutonium in nearby sediment.

What Remains Genuinely Unresolved

None of that settles the matter permanently, because none of it is a one-time finding — it is a snapshot of a structure that keeps corroding. The sealant and titanium plugs fitted to the bow in the mid-1990s were engineered for an estimated 20-to-30-year service life, a window that is closing now or has already closed. Strontium-90, also present in reactor fuel, was flagged by the 2019 report for further laboratory analysis but has not been confirmed as currently escaping. And joint Norwegian-Russian monitoring, the only mechanism producing independently gathered data on the wreck, ceased in 2022 amid broader diplomatic rupture — leaving the most consequential open questions about this site answered, for now, only by whoever Russia allows near it.

The Bottom Line for an Evergreen Risk

The honest summary is neither alarmist nor dismissive. The reactor leak is real, confirmed, and recurring across three separate observation periods spanning thirty years. The health and ecological risk from that leak, as measured so far, is negligible because of rapid dilution at depth — a conclusion shared by the Norwegian scientists who found the contamination in the first place. The plutonium scenario remains the wreck’s true long-tail risk: unconfirmed today, dependent on aging patches nobody is currently verifying, and effectively unmonitored since cooperation broke down. That combination — a confirmed small problem sitting beside an unconfirmed large one, under a custodian with every incentive to minimize both — is why Komsomolets will remain a live concern for as long as it sits on the seabed.