Set some 50 years in the future, every spaceship in space is powered by a nuclear powered reactor and most of them came equipped with an ion propulsion engine for acceleration. I noticed that every spaceship's captain and engineers are drilled to wait for at least 60 hours before restarting the nuclear reaction again, especially after their ship had experienced a power phase out; it is clearly stated in the protocol and there will be a hefty fine for those who blatantly ignore it. I heard from some of those who has retired from the service mentioned that it could be to prevent nuclear poisoning, but come on, we are already in the space age and we still need to wait for the spaceship's engine to warm up? Why is that? any solution to expedite this?

  • 9
    $\begingroup$ The reason I did not answer was that it seemed to me that the OP knows about neutron poisons, but for some reasons wants a different explanation. So I asked for clarification, if this is indeed the case. $\endgroup$
    – b.Lorenz
    Jun 13, 2020 at 12:43
  • 4
    $\begingroup$ You tag science-based, but you're both handwaving the actual physics and introducing a non-sequitur in your question. $\endgroup$ Jun 14, 2020 at 5:00

3 Answers 3


Your coolant takes time to melt

Back in the days of the Soviet Union, there was a class of high-performance submarine called the Alfa class.

This submarine used a lead-cooled reactor. By doing so, it managed to compact the size of the reactor significantly, create enormous thrust allowing the submarine to even outrun torpedoes.

This caused the U.S. lots of anxiety, until it was realised the downside of such reactors was that the lead must be kept constantly in a liquid (molten) form and if the reactor stopped running, it would solidify and create a huge chunk of lead in the reactor - permanently disabling the submarine.

As a result, maintenance of the submarine was a major problem - for simplicity the reactor was kept on an 'always-on' mode, however for large periods of downtime there were significant problems with heating and pumping this type of coolant at coastal facilities.

Your spaceships could have similar design characteristics to the Alfa class, requiring any of the following:

  • The need for high-performance reactor
  • The need to keep reactor space minimal (easy to imagine in a cramped spaceship)
  • The inability to keep a reactor 'always-on' and requiring coolant pumping and heating

Such constraints are easily conceivable in a spaceship with design trade-offs required to offset issues of power, size and weight.

  • 13
    $\begingroup$ On realistic spaceships, volume is rarely as strong a constraint as mass. If these reactors are also less massive than the alternative, then that would be a good point to add. $\endgroup$
    – BBeast
    Jun 13, 2020 at 10:17
  • 3
    $\begingroup$ This goes straight to becoming railworld canon :D $\endgroup$
    – dot_Sp0T
    Jun 13, 2020 at 13:01
  • 5
    $\begingroup$ I know that salt reactors have been gaining popularity lately. Since they are also solid-cooled, do they have the same issue? $\endgroup$
    – Nosajimiki
    Jun 13, 2020 at 13:31
  • 1
    $\begingroup$ @BBeast - yes actually for an equivalent Power, this reactor does have less mass as it is more efficient and has a higher thermo-dynamic range to generate power. I suppose though that if your spaceship needed to keep a very low aspect ratio to either a target, or say ablative effects of interstellar dust, perhaps minimising size also is beneficial in space. $\endgroup$
    – flox
    Jun 14, 2020 at 2:29
  • 1
    $\begingroup$ I wonder if you could build a ship with two reactors sharing one cooling system, such that only one can be on at a time, but you can always keep your coolant liquid by shunting it to the other reactor while you take the first offline to do maintenance. Probably expensive, but maybe answers the "expedite" part of the question? $\endgroup$
    – kojiro
    Jun 14, 2020 at 16:48

XENON Poisioning

A fission reactor works by maintaining a delicate and precise balance between creating and destroying free neutrons. When U235 splits, it releases neutrons which can hit other U235 nuclei and cause them to split.

If the growth rate of that process is greater than one, you have an explosion. If less than one, the process fizzles. If exactly equal to one, the reaction is sustained.

Simple, really.

But, the reaction environment does not stay constant.

The U235 splits into other elements with varying half-lives and nuclear behavior. Xenon-135 is short-lived isotope produced from the decay of another fission product. Xenon-135 is an especially good absorber of neutrons, which makes it hard to get a sustained reaction. Doing so requires removing much of the deliberate neutron absorbing control rods, and makes the reaction, if it gets started, difficult to control. Here is a video explaining the problem, and the Wikipedia page: Xenon Poisioning.

It is a real problem, and could result in a loss-of-ship event.

A little more data inspired by comments

Xenon-135 is not formed immediately when Uranium is split. It is formed when Iodine-235 decays into Xenon-135. The half-life of Iodine-135 is short (Wikipedia reference), but because of that half-life the level of Xenon-135 takes a few hours to stabilize at new level. This creates a complex damping effect as a reactor's power (and thus the rate of fuel fission) increases, and instability as power is decreased.

Xenon-135 probably played a role in the Chernobyl melt-down and fire. The output of the reactor had been substantially reduced for a few hours, and then was ordered to be increased. The Xenon-135 load was higher that steady state for the reduced power level. To increase the output, the influence of the control rods needed to be reduced. It is likely that a localized reaction then grew out of control and ignited the graphite moderator.

This problem exists in any fission reactor that produces Iodine-135, and is not limited to a U-235 reactor. Thorium and Plutonium reactors must consider the levels of Xenon-135 in the control practices.

  • 6
    $\begingroup$ You could add that this was a process that was one of the causes of the chernobyl reactor disaster, and that this is basically the main, physical requirement in uranium based reactors, so as long as the reactor is a uranium reactor, it will ALWAYS need this kind of delay. $\endgroup$
    – Gensys LTD
    Jun 13, 2020 at 19:36
  • 2
    $\begingroup$ The problem is not limited to U-235 fuel. Plutonuim-239, MOX, Thorium/U233 or any other neutron-mediated fission chain reaction will have the same iodine-xenon "pit". "Molten salt" reactors are able to manage the problem, to an extent - removing Xe-135 from the reactor instead of burning it produces more of the unwanted Cs-135 isotope that is harder to dispose of later. $\endgroup$
    – fraxinus
    Jun 13, 2020 at 20:21
  • 1
    $\begingroup$ Also somewhat related: buildup of Samarium 149 might lead to inability to restart reactor without replacing fuel because it does not decay naturally. $\endgroup$ Jun 14, 2020 at 15:20
  • $\begingroup$ FYI - I don't think the link to the video points exactly where you mean it to. $\endgroup$
    – skifans
    Jun 15, 2020 at 13:23
  • 1
    $\begingroup$ @Skifans, thank you. You were 100% correct. $\endgroup$
    – cmm
    Jun 15, 2020 at 20:43

Because the reactor is sealed and inaccessible to crew whether operating or not, the maintenance cycle for a restart requires the use of dozens of tiny robotic drones to crawl through the pipeways and ductwork looking for and repairing minor cracks and faults. Although the chance that the damage to a reactor from a single phasedown would be so bad as to cause disastrous consequences without that maintenance pass is small, it IS nonzero, and it's definitely true that as maintenance cycles are skipped, the chance of a catastrophe skyrockets.

Thus, all prudent shipmasters allow the drones the full 60 hours that it typically takes to run a comprehensive cycle.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .