It's possible that reliability/safety will improve as fast as the launch cadence does, in which case we could see the accident rate staying flat or decreasing.

Launches are less to be feared because they are oriented toward sea and the most critical moment are the first minutes. The device discussed is intended to travel away from earth so as you noted it won’t be dangerous after it quit orbit.

Disclaimer: not skilled in any related field, just a consumer of space science popularization videos for newbies.

(Not a direct comment on your opinion specifically, just building of the general idea)

Oceans are the cradle of life and are still populated by fishes, mammals, zoo and phytoplancton even away from coastlines.

http://comlmaps.org/gallery/footprints/

https://www.rnz.co.nz/international/programmes/datelinepacif...

Abyssal zones of the ocean are only populated by creatures that feed on hydrothermal vents, or dead creatures falling down from the surface. But the mid pacific has no hydrothermal vents and the surface is a dead zone due to human activity. So there is no source of food for life to exist here, and outside of the occasional whale fall it is truly dead.

I'm kinda surprised the soviets even thought of the environment in that way.

Consequently, reactor payloads would be carried on a very small fraction of launches, then used in space for an extended amount of time.

    > Or even if it broke up
    > in the atmosphere.

It’s not the 99 rockets falling into the ocean shortly after launch that are problematic, it’s the 1 that almost works but then lands in the city we should be concerned with.

That's pretty close for what would be high level nuclear material.

A better way would just be shielding, make sure it stays in 1 piece through reentry and impact. And still doesn't leak radiation then.

Also you can presumably shield reactors to survive reentry or even ship the fuel in multiple such reentry proof containers. And of course longer term build these in space, given nuclear propulsion is basically in space only anyway given its characteristics.

Or it could go the way of the airline industry where failures become less regular as the industry develops ways of systematically reducing them.

Also, nuclear rockets should probably be launched to earth orbit in pieces, separately, so that if any one launch fails it won’t create a nuclear accident in the biosphere. Then assemble and power up the nuclear rocket as far away from earth as possible.