Nuclear power question

[AllFreedomUnlessDefyingScience]

Say you want to send a probe into space with a working nuclear reactor that produces power for as long as possible (assume you have already solved the massive problems involved in automating this process). If a specific example of size and output is needed assume we have a modern nuclear power plant for civilian power generation (I'm not sure of the exact parameters so assume something in the middle of the road). Nuclear fuel has a decay rate. How fast would you run out of fuel for your nuclear reactor? Can anyone offer a response, or is my question fundamentally flawed?

Thanks for any help,
AllFree

[pericynthion]

Tens of years. Submarines regularly go 20+ years without refueling. If you dial down the power output, you could probably get at least 100 years out of it. There'll be a certain minimum power level to run housekeeping stuff for the reactor, so it can't be indefinite.

[AllFreedomUnlessDefyingScience]

Because of the nature of nuclear decay, I imagine you could carry larger amounts of nuclear fuel, planning on most of it to decay before you use it, however this would hugely increase the mass of the ship and quickly reach the point of diminishing returns correct?

[pericynthion]

Fission fuel (U-235 or Pu-239 in MOX) has such a long half-life that for most purposes you can consider it not to decay at all. The problem with carrying a core overstuffed with fuel is similar to cramming a wood stove totally full - the ash (fission products) build up and poison the reaction.

In principle if you needed a very long operating life you could store extra fuel outside the core and build an automated refueling apparatus, but that's way more complicated, heavy and failure prone than keeping all the fuel in the core.

[AllFreedomUnlessDefyingScience]

Originally Posted by pericynthion

In principle if you needed a very long operating life you could store extra fuel outside the core and build an automated refueling apparatus, but that's way more complicated, heavy and failure prone than keeping all the fuel in the core.

Since I've already assumed an unrealistic state of automation for the reactor, lets further stipulate to the same state for external storage. If this was the case, what kind of times are you talking about reducing the decay rate too?

[Der Trihs]

Maybe a traveling wave reactor would be a good idea for this (although one has yet to be actually built)? According to the article it should run for 50-100 years, without refueling or having to remove waste. It starts with all the fuel already in the core.

[Quartz]

Didn't the Soviet Union routinely send reactors into orbit?

[Xema]

Originally Posted by Quartz

Didn't the Soviet Union routinely send reactors into orbit?

They - along with the US - have sent substantial numbers of RTGs (radioisotope thermoelectric generators) into space:

A common application of RTGs is as power sources on spacecraft. Systems Nuclear Auxiliary Power Program (SNAP) units were used especially for probes that travel far enough from the Sun that solar panels are no longer viable. As such they are used with Pioneer 10, Pioneer 11, Voyager 1, Voyager 2, Galileo, Ulysses, Cassini and New Horizons. In addition, RTGs were used to power the two Viking landers and for the scientific experiments left on the Moon by the crews of Apollo 12 through 17 (SNAP 27s). Because Apollo 13 was aborted, its RTG now rests in the South Pacific ocean, in the vicinity of the Tonga Trench. RTGs were also used for the Nimbus, Transit and LES satellites.

As for true reactors:

By comparison, only a few space vehicles have been launched using full-fledged nuclear reactors: the Soviet RORSAT series and the American SNAP-10A.

[engineer_comp_geek]

Space probes tend to use thermoelectric generators, as Xema mentioned. These are dirt simple power generators, and because of their simplicity, they tend to be very reliable and last for a very long time. Basically, they just use some sort of radioactive material to generate heat, which is then used to heat thermocouples or thermoelectric generators (peltier devices) to create electricity. There are no moving parts. It's not the most efficient way of generating electricity, but for something like a space probe you aren't going to find anything that is more compact, rugged, reliable, and long lasting.

The power output is going to drop over time. If you design your probe to use only a tiny fraction of the power you generate, you could easily make your probe last for 50 to 100 years. Getting the other electronics in the probe to last that long is going to be a significant challenge.

By comparison, a typical "middle of the road" civilian nuke plant these days is a monstrously huge and complex beast. It is designed to be a major workhorse in the power industry, and is going to generate 1000 MW or more. I can't imagine a probe that would use anywhere close to that amount of power. If you want this sort of traditional type of nuclear reactor, you'd do much better with something much more similar in size to a nuclear sub's reactor. Making one that is reliable enough to last a hundred years without anyone doing any maintenance on it is going to be very tricky.

[JWT Kottekoe]

As stated above, you needn't worry about the decay of the fissile material. The half life of Uranium 235 is about 700 million years, Plutonium 239 has a half life of 24 thousand years.

To get seriously mass efficient use of the nuclear material, you need to go with something like Project Orion, blowing off hydrogen bombs to achieve enormous thrust and specific impulse. Don't try this at home.

[Polycarp]

One point not yet made is that what you want is a nuclear power source -- though in the OP you specified that it be a reactor. But a radioisotope with a half-life in the tens or hundreds of years is going to produce atomic power consistently (though at a gradually declining rate) for a time span on the order of a millennium. You could easily use high-level nuclear waste for this -- in fact, the very properties that made it so difficult to dispose of safely are the ones making it useful in a scenario like this.

[JoelUpchurch]

All space based nuclear reactors have a problem dissipating waste heat, which limits their output. The reactors end up being mostly radiator fins and even then the power output is pretty pathetic.

Compare this picture of the Snap 10A

With this picture of a 8DR core

On the Moon or Mars there are more ways to dissipate heat and they are also using more efficient Stirling engines to design reactors to generate up to 40KWe. It looks like it still needs a lot of radiating surface. The article below describes over 1000 square feet for the final system.

http://www.technologyreview.com/energy/23247/?a=f

[CutterJohn]

Originally Posted by engineer_comp_geek

By comparison, a typical "middle of the road" civilian nuke plant these days is a monstrously huge and complex beast. It is designed to be a major workhorse in the power industry, and is going to generate 1000 MW or more. I can't imagine a probe that would use anywhere close to that amount of power. If you want this sort of traditional type of nuclear reactor, you'd do much better with something much more similar in size to a nuclear sub's reactor. Making one that is reliable enough to last a hundred years without anyone doing any maintenance on it is going to be very tricky.

That monstrous complexity comes from the mass of redundant systems, safety systems, and systems designed to eke every last bit of efficiency out of the system.

All a reactor really needs is 1 control rod mechanism, 1 turbine, and 1 pump. Oh, and a few gauges for the control systems. Automating would be pretty easy.

Still, they are moving parts. Making it last a long, long time will, as you say, be tricky. Making it last merely a long time... Not so bad. The primary systems I worked with on the enterprise were built absurdly tough, and lasted a good long time. If only the secondaries had been built like that...

[Quartz]

Sounds like a nuclear reactor would be an ideal power plant for a mobile probe for Titan, what with a nice thick atmosphere with which to dissipate the heat.

[JoelUpchurch]

Originally Posted by Quartz

Sounds like a nuclear reactor would be an ideal power plant for a mobile probe for Titan, what with a nice thick atmosphere with which to dissipate the heat.

Titan would be a good candidate or Europa, which is thought to have a water ocean under the ice. Currently Europa is the most interesting object in the solar system, since scientists thing there might be life down in the ocean there. I suppose you could also melt the ice to carry the heat away.

From a commercial point of view, the moon or an asteroid would be more valuable for supplying raw material for orbital industries.

[enigmatic]

I'm a little reluctant to let go of this idea, cause it forms a key plot point on the sci-fi setting I've been polishing up in my head. But instead of using the standard sci-fi idea of a spacecraft somehow carrying a small amount of antimatter as a fuel source, you construct the spacecraft out of antimatter and figure out a design that would allow it to use background hydrogen as a fuel source (easier said than done for one obvious reason ).

Unless I'm missing something, once it is travelling through a vacuum the fact that it's composed of antimatter has little downside unless it hits something big in which case it was probably doomed anyway. I'm envisaging a very small light probe here, less than 100 g or so, otherwise the antimatter production cost would seem prohibitive even in fiction.

[FoieGrasIsEvil]

Originally Posted by engineer_comp_geek

Space probes tend to use thermoelectric generators, as Xema mentioned. These are dirt simple power generators, and because of their simplicity, they tend to be very reliable and last for a very long time. Basically, they just use some sort of radioactive material to generate heat, which is then used to heat thermocouples or thermoelectric generators (peltier devices) to create electricity. There are no moving parts. It's not the most efficient way of generating electricity, but for something like a space probe you aren't going to find anything that is more compact, rugged, reliable, and long lasting.

The power output is going to drop over time. If you design your probe to use only a tiny fraction of the power you generate, you could easily make your probe last for 50 to 100 years. Getting the other electronics in the probe to last that long is going to be a significant challenge.

By comparison, a typical "middle of the road" civilian nuke plant these days is a monstrously huge and complex beast. It is designed to be a major workhorse in the power industry, and is going to generate 1000 MW or more. I can't imagine a probe that would use anywhere close to that amount of power. If you want this sort of traditional type of nuclear reactor, you'd do much better with something much more similar in size to a nuclear sub's reactor. Making one that is reliable enough to last a hundred years without anyone doing any maintenance on it is going to be very tricky.

What kind of maintenance is performed on nuclear submarine reactors? How big/heavy are they?

[Chronos]

Currently Europa is the most interesting object in the solar system, since scientists thing there might be life down in the ocean there.

Nitpick: Second-most interesting.

[robby]

Originally Posted by FoieGrasIsEvil

What kind of maintenance is performed on nuclear submarine reactors?

A lot. There's daily, weekly, monthly, quarterly, and annual preventive maintenance, along with major overhauls every 3-5 years and refuelings every 15-20 years.

Originally Posted by FoieGrasIsEvil

How big/heavy are they?

They're relatively small when compared to a civilian reactor, but still much bigger than something you could put on a spacecraft. They're also quite massive, primarily because of the shielding.

A naval design is not something you could easily adapt for a spacecraft. There are simply too many moving parts and too much maintenance required, and the naval reactor is mated to a steam plant, which is not the most practical thing for a spacecraft. Removal of waste heat becomes a huge issue without the heat sink of an ocean, so you'd need a way to radiate off the heat with radiator fins.

This is why the RTGs (radioisotope thermoelectric generators) mentioned upthread generally make more sense for a spacecraft.

[Quartz]

Originally Posted by enigmatic

I'm a little reluctant to let go of this idea, cause it forms a key plot point on the sci-fi setting I've been polishing up in my head. But instead of using the standard sci-fi idea of a spacecraft somehow carrying a small amount of antimatter as a fuel source, you construct the spacecraft out of antimatter and figure out a design that would allow it to use background hydrogen as a fuel source (easier said than done for one obvious reason ).

Anti-matter? This is probably your one obvious reason, but space isn't that empty. Interplanetary space has a density of 5-10 protons per cc from the solar wind. Never mind grazing a planet's exosphere or a comet's tail. That doesn't sound a lot, but you're going to be travelling very fast and going to hit an awful lot of them and their annihilation will be very energetic. It would swiftly be eroded by the solar wind and micrometeorites.

[enigmatic]

Anti-matter? This is probably your one obvious reason, but space isn't that empty. Interplanetary space has a density of 5-10 protons per cc from the solar wind. Never mind grazing a planet's exosphere or a comet's tail. That doesn't sound a lot, but you're going to be travelling very fast and going to hit an awful lot of them and their annihilation will be very energetic. It would swiftly be eroded by the solar wind and micrometeorites.

Partly, the obvious problem was figuring out a plausible design that could use collisions occurring primarily on the front of the craft for propulsion. I was certainly assuming a much lower background matter density (I was thinking a similar number per square meter or so rather than cc, and probably guesstimating less energy loss with each collision, I haven't done any actual math ). This is all much less of a problem for how I was actually intending to use the basic idea in the story, the idea of using it in a remotely plausible, hard sci-fi ish space probe was a separate one.