[Enthalpy]

Explosion and reported brief radioactive contamination at a Russian missile test facility in Nyonoksa, 47km from Severodvinsk
https://www.bbc.com/news/world-europe-49301438
https://thebarentsobserver.com/en/security/2019/08/latest-rosatom-says-five-employees-killed-blast-while-testing-isotope-and-liquid
https://www.france24.com/en/20190810-russia-says-five-died-missile-test-explosion
https://www.rt.com/russia/466194-rosatom-killed-arkhangelsk-blast/
https://www.thedrive.com/the-war-zone/29356/russia-admits-mysterious-missile-engine-explosion-involved-nuclear-isotope-power-source
the newspapers of varied origins and quality seem to have very few sources, Rosatom and the Defence ministry, plus unverified videos. 5 dead, 3 wounded.

The cruise missile is supposedly a Burevestnik, said to be propelled by a nuclear reactor for big range, as shall the Poseidon torpedo be
https://en.wikipedia.org/wiki/9M730_Burevestnik
https://en.wikipedia.org/wiki/Status-6_Oceanic_Multipurpose_System
https://www.youtube.com/watch?v=okS76WHh6FI

but the reports on the explosion tell instead
"liquid-propellant engine"
"isotope power source" for a propulsion system
which isn't necessarily a nuclear reactor, where fission happens.

One source fitting the wording would be a big amount of natural or artificial short-lived isotopes that heat the air that passes by in a ramjet, or hydrogen from a tank. This needs a huge radioactivity obtained from a big reactor few hours to weeks before flight. Not convenient.

Or maybe isotopes only heat the propellant before it burns with air, in a bizarre pumping cycle for ramjet or scramjet
https://www.chemicalforums.com/index.php?topic=91121.msg333349#msg333349 and next ones
but I see essentially drawbacks to that. While pumping for a scramjet isn't trivial, chemical or thermal paths bring less worries.

Explanations, suggestions...?

[Enthalpy]

Trying to put hard figures on some bizarre assertions...

Let's imagine a missile propelled by a radioisotopic souce. Not even a hypersonic scramjet, just an easy subsonic cruise missile with extended range.

I take 20t mass, flying low at 250m/s = Mach 0.73, with L/D=5 as cruise missiles resemble a V1 more than a glider. The small engine must push 40kN and 10MW, for which it needs >40MW heat, more in a ramjet.

A big nuclear reactor running at full 4.2GW heat output is full of radioisotopes. I take the fission products ⁹⁹Mo as an example.
https://en.wikipedia.org/wiki/Fission_product_yield
https://en.wikipedia.org/wiki/Isotopes_of_molybdenum
If it had equilibrium concetration in the reactor, is separated and used immediately, its 6.05% fission yield and mean 389keV beta energy
https://www-nds.iaea.org/relnsd/vcharthtml/VChartHTML.html
https://www.nndc.bnl.gov/nudat2/reCenter.jsp?z=42&n=57
https://www.nndc.bnl.gov/nudat2/decaysearchdirect.jsp?nuc=99MO&unc=nds
compared with mean 200MeV for uranium fission let this isotope produce at most 500kW heat. Its daughter ^99mTc provides a little additional energy as gammas. Keeping half a dozen main fission products lets reach 3MW heat, not >40MW. Some leave more decay time than ⁹⁹Mo but are heavier for the same power. The user would need to stop 10 big nuclear reactors, process and bring the waste quickly to the weapon waiting for use.

Impossibility can't be proven in technology, and proofs belong to mathematics only, but under the very broad assumptions above, I comfortably claim that radioisotopic power doesn't propel a missile. It's not only impractical: the figures don't fit.

==========

But a nuclear reactor could propel that cruise missile. Reactors can be powerful and light, as Nerva showed
https://en.wikipedia.org/wiki/NERVA
18t for the whole engine, where the reactor produced 1GW. An engine heating air instead of hydrogen would push more than Nerva's 246kN. The mass of a nuclear reactor doesn't scale like its power, but the 20t missile looks feasible, or a bigger one, and it can fly for long and faster, like Mach 2 or 3.

Hypersonic speed is more doubtful. If ejecting hydrogen from a tank, 800s specific impulse and L/D=3 let fly for <<40min or <<5000km, rather 4min and 500km if bulky hydrogen makes 10% of the start mass. Long flight demands to heat air in a ramjet, despite compression from high speed heats it already. It's conceivable with low efficiency as in a scramjet if the flow remains supersonic hence lukewarm before the heater which is then quite difficult. Or the reactor must heat the air by the emitted neutrons rather than by contact with reactor parts that are cooled below the ramjet's chamber temperature. Though, neutrons carry a minor part of the fission energy, so cooling the reactor takes a bigger air flux than the ramjet has, but with less drag than the ramjet pushes. Badly difficult too.

If such weapons are more dangerous for the shooter than for the targets, it's good news for the targets.

[Enthalpy]

Since the thing that exploded in Nyonoksa was tested on a platform at sea rather than on the strand, reasonably it should be an aquatic weapon, not a missile. Is an isotopic power source more meaningful then?

Abundant ⁹⁰Sr shall be the radioelement, as Russia often chooses it against the safer ²³⁸Pu and ¹⁴⁷Pm. The badly dangerous Bremsstrahlung of daughter ⁹⁰Y would explain the radiation accident. The compounds aren't volatile, but accidents at nuclear reactors disseminate ⁹⁰Sr, so maybe a big uncooled radioisotopic source does too, and the detected radioactivity was small anyway.

The fission yield is 4.5% so one big reactor producing mean 3200MW heat a year long accumulates 233 moles that weigh 43kg as SrTiO₃. Several reactors over several years would accumulate more, but there would be several weapons too. Including ⁹⁰Y, the activity is terrifying 213PBq whose Bremsstrahlung can't be fully shielded. Mean 196+934=1130keV beta energy produces only 39kW heat convertible to some 10kW propulsion power. similar sources powered lighthouses, needing a comparable power, and sufficed to warm hunters sleeping nearby.

A torpedo with ⁹⁰Sr propulsion and small nuclear warhead wouldn't resemble the published picture. It could have D=0.5m for a 100kt hydrogen bomb that destroys a naval group, and L~6m. To help the weak propulsion, it would be streamlined not for C_d=0.08 like a submarine or a sausage, but rather for C_d=0.02, or possibly better.

Even so, 10kW achieve only 17m/s = 33 knots. That's as slow as the targeted naval group, bad for a torpedo.

A radioisotopic source gives range. 30 000 km to the antipodes with detours take 20 days while ⁹⁰Sr provides power for >10 years. If the war isn't over, and a suitable target like a port can be found, and the torpedo controlled, it is less detectable than a submarine and resembles more a fish. But other power sources do it better.

Idling mines or torpedoes are more meaningful. ⁹⁰Sr would let them rest on the deep Ocean bottom or navigate slowly for decades. Hypothetical communications would wake one weapon up to attack a naval group passing nearby. They could use chemical propulsion (accident reports tell "liquid propulsion") near the target. The radioisotopic source could be less huge.

That would be an answer to the shallow Barents and Baltic seas, where submarines are easily detected and Russia has naval bases.

[Enthalpy]

New test of Burevestnik 9M370 (SSC-X-9 Skyfall for Nato) this month, reportedly successful for the first time.

Consistent part of the description, allegedly provided by Russia:

• Propelled by a nuclear reactor for practically unlimited range and duration.
• Subsonic flight at low altitude with maneuvers. A cruise missile hence.

The rest appears to be blah-blah: undetectable, invincible and so on.

If Burevestnik boasts nothing else, its targets can sleep equally well. Pollution is the only added risk. Such a cruise missile is perfectly detected at a huge range and any altitude by over-the-horizon radars OTH
  Burevestnik - over-the-horizon at Wiki
After detection, any combat aircraft can down the thingy.

OTH observe at arbitrarily low altitude and distances like 3 000 km by reflections at the ionosphere, while their decametric waves are nearly unhampered by anti-radar design. Old stuff, but we may need more of them, like at the tip of Norway and in Turkey.

Bad luck for the "unexpected direction of attack": the USA have already OTH covering the Caribbean and much of South America.

I suspect the deployment of OTH stopped because more convenient AWACS bring similar capabilities.

Most technological countries also use the radiation of civilian TV and radio transmitters measured by military networks of receivers to synthesize huge radar arrays over a wide frequency range, but this is less known.

[marquis]

There are websites, such as kiwisdr, that allow listening to the HF ( highfrequency) radio. A lot the OTH ( over the horizon) radars are in that area.  And while the data produced is well above the audio rate, it seems well below what would be produced by something like an AWARd radar. Just a guess...Time in to a kiwisd r sites and tune one of the OTH sits, you will see.  The exact frequency depends on propagation.  But, if you look it up on Wiki, the sound is pretty identifiable.  Thank you for the other information on the nuclear jets.  . Regards