The natural fit here is a naval nuclear reactor. ... Of course, this raises some thorny issues around potential launch failures spreading radioactive material o

The natural fit here is a naval nuclear reactor. ... Of course, this raises some thorny issues around potential launch failures spreading radioactive material over a large area. This is already a consideration even with the little RTGs that NASA launches for rovers. Counter-intuitively, a nuclear reactor has lower radiotoxicity risks in case of an explosion during launch. Before a reactor reaches criticality for the first time, it doesn't contain any highly radioactive materials. It's only as radioactive as its uranium fuel. Uranium has a long half-life and corresponding low radioactivity. Plutonium 238, used in RTGs like that in the Mars Science Laboratory, has a half life of 87.7 years. Uranium 235 has a half life of 704 million years. Neglecting differing relative biological effectiveness of radiation for the moment, that means that the plutonium 238 in RTGs is about 8 million times more radioactive, kilogram-for-kilogram, than the uranium 235 in a reactor's fuel. The Mars Science Laboratory carried 4.2 kilograms of plutonium 238 (4.8 kg of PuO2): https://www.nasa.gov/pdf/532433main_FIN%20MSL%20Launch%20Nuclear%20Safety%20FS%202-9-11.pdf https://www.nasa.gov/pdf/532433main_FIN%20MSL%20Launch%20Nuc... It could have carried 33.7 million kilograms of U-235 for roughly the same radiological risk. Or, realistically, missions could launch with reactors designed for a few tens of kilowatts to megawatts and the maximum radiological risk would be much smaller than with RTGs. At least so long as the reactor attains first criticality after the risky launch phase.