Will SpaceX ever go nuclear propulsion?

[Nilof]

If you mean the fission products, some LFTR reactor designs have a filtering mechanism where these are removed from the reactor. This is one MAJOR advantage of liquid fuel, since this is impossible in solid-fueled reactors as the products are trapped inside the reactor and can cause neutron poisoning.

Nope, this incorrectly suggests fission products generally are filtered out. Xenon-135 is removed as a neutron poison, and I suppose there's other gasses that go with it, but most of them stay in the salt and are quite capable of causing significant decay heat after shutdown.

It also means that in the case of an actual accident/containment vessel breach, there is no built-up radioactive material that can escape.

Unequivocally false.

Either a FP boils off easily and is easy to remove, or it doesn't and thus can't escape easily from the fuel salt in case of an accident. In current reactors, a meltdown is dangerous in the first place precisely because built up fission products like Cesium-137 can boil off and escape from the melt.

The fuel salt is temperature self-regulating, so uncontrolled salt will not be hotter than in routine operation. So any elements that could boil off and be released would already boil off in routine operation. This means that the only radioisotopes left would be stuff like Strontium-90 which have a too high boiling point to escape, and indeed Sr-90 has not been a problem during previous catastrohpic failures such as Chernobyl even though it was a majority fission product.

[macpacheco]

...okay, now imagine getting all that kind of safety in any kind of flight-weight system. This is why SpaceX will never go with nuclear propulsion. That weight is not a big deal on Earth, very big deal in space.

I wouldn't think like that. Salt coolants have so much higher thermal capacity per Kg than water and the atmospheric pressure feature means internal piping is much thinner, plus a water cooled reactor with a proper terrestrial secondary containment would weigh an absurd weight, so an MSR with all the safety features is like less than 1/3rd of the weight and volume than a similar terrestrial reactor per MWe.
So on the weight and volume budget, MSRs are way ahead LWR reactors for terrestrial usage. But I have no clue what safety features will be absolutely required and which ones wouldn't for launch vehicles or spaceships. Outside the atmosphere we usually have ultra low temperature sources (vacuum, assuming it can always be away from the Sun), so cooling might be much easier than on earth, which further reduces demands on safety features.

[Robotbeat]

Cooling is WAY harder in space because you're operating via radiative cooling instead of convective/conductive. Not so bad on Mars, but it's still harder than dumping heat into a river. In space, you need a big, heavy radiator, especially if you claim to operate at high efficiency.

But regardless, most space reactor designs use helium or something like that for coolant, not water. I don't know if ANY proposals I've read use water, in fact.

[ArbitraryConstant]

Either a FP boils off easily and is easy to remove, or it doesn't and thus can't escape easily from the fuel salt in case of an accident. In current reactors, a meltdown is dangerous in the first place precisely because built up fission products like Cesium-137 can boil off and escape from the melt.

Don't believe the helium sparge has the effect of keeping Cesium-137 from building up. For the Xenon-135, the main neutron poison is serendipitously a noble gas that will bubble out if you ask it nicely, most fission products do not afford you this luxury.

The fuel salt is temperature self-regulating, so uncontrolled salt will not be hotter than in routine operation.

Confused by this, seems incorrect. The fission reaction can have a temperature coefficient of reactivity, but not decay heat. That happens according to half lives of the isotopes, nothing can be done about it other than try to remove the heat by safe means before it removes itself by unsafe means.

This means that the only radioisotopes left would be stuff like Strontium-90

Also wrong, you're talking about long lived fission products. Decay heat of a freshly shut down reactor is dominated by short lived isotopes.

[IslandPlaya]

I must find a way to introduce your fine word 'sparge' in casual conversation... 

[Avron]

I must find a way to introduce your fine word 'sparge' in casual conversation... 

"A step in lautering, a process used in brewing beer'  -- now we are talking

[JasonAW3]

This wasn't the first time that the military conservativeness did us a sizeable diservice picking inferior technology.

Unsubstaniated

When you consider part of the reason that the military chose the water cooled reactors, not just for simplicity but to also partially process radioactive materials into materials bthat could be used in nuclear weapons, using a liquid metal reactor makes little sense.

But what the previous commenter failed to realize is that the vast majority of nuclear powerplants were to be used around water, the use of liquid sodium reactors is damned near suicidal!  Sodium in water is a bad thing as sodium explodes in water.
Moltent, liquid sodium, exposed to excessive amounts of sea water would cause an explosion and a potential nuclear accident of such a degree that Chernobyl would seem like a minor microwave malfunction.

I have to agree with Jim on this one.  Between nuclear incidents that have been declassified and ones that will likely remain classified forbthe rest of our lives, the choice that the military has made in the type of nuclear power system that the currently use, is the best and safest choice that could be made with the technology of the beginning of the nuclear age.

[macpacheco]

This wasn't the first time that the military conservativeness did us a sizeable diservice picking inferior technology.

Unsubstaniated

When you consider part of the reason that the military chose the water cooled reactors, not just for simplicity but to also partially process radioactive materials into materials bthat could be used in nuclear weapons, using a liquid metal reactor makes little sense.

But what the previous commenter failed to realize is that the vast majority of nuclear powerplants were to be used around water, the use of liquid sodium reactors is damned near suicidal!  Sodium in water is a bad thing as sodium explodes in water.
Moltent, liquid sodium, exposed to excessive amounts of sea water would cause an explosion and a potential nuclear accident of such a degree that Chernobyl would seem like a minor microwave malfunction.

I have to agree with Jim on this one.  Between nuclear incidents that have been declassified and ones that will likely remain classified forbthe rest of our lives, the choice that the military has made in the type of nuclear power system that the currently use, is the best and safest choice that could be made with the technology of the beginning of the nuclear age.

Most reactors produce lots of plutonium, water cooled, gas cooled, but they produce very impure plutonium (aka reactor grade plutonium, contaminated with Pu-240 and Pu-241). Even fast reactors and MSR reactors produce plutonium too.

Nuclear weapons are built with weapons grade plutonium produced on low power thermal spectrum reactors where a blanket of depleted Uranium (99.9+% U-238) gets slowly irradiated with neutrons (then plutonium is separated from Uranium) or simply highly enriched uranium (80% or higher U-235).

While the majority of operational nuclear weapons were produced with plutonium, it is now much more economical (and technically easier) to just make richly enriched uranium. Notice people aren't talking about Iran making plutonium nuclear weapons, instead the problem is highly enriched uranium.

Some operational commercial reactors could produce weapons grade plutonium, but it would involve a depleted uranium blanket as well, in a low neutron flux area of the reactor. You must minimize U-238 getting two neutrons (one makes Pu-239, another makes Pu-240, aka the bad stuff).

I'm unaware of operational nuclear weapons made with plutonium produced from spent nuclear fuel, as it would require first separation of plutonium, then enrichment of plutonium. SNF comes from the core of the reactor instead of the blanket. Its way easier to separate U-235 from U-238 (3 neutrons in weight difference) vs Pu-239 from Pu-240 (a single neutron in weight difference).

This is a very common misconception. The same reason that U-233 is a bad nuclear weapons fuel, as it gets contaminated with U-232 that suffers from high energy gamma rays that offers the same risks (premature detonation and degradation of the high explosive primary and degradation of the control electronics of the weapon).

As far as sodium powered reactors inside a nuclear sub, molten salt reactors running on Uranium were a possibility, they were abandoned for political reasons (MSRs were primarily considered using Thorium and since Thorium produces U-233 that is bad nuclear weapons material, it was considered the bad way, but a Thorium+U-233 reactor could be used to irradiate a U-238 blanket and make weapons grade plutonium as well).

I guess my biggest criticism is towards commercial nuclear power using water cooled reactors. All water cooled reactors requires maximum active safety systems compared with fast sodium and MSR reactors. Three Mile Island, Chernobyl and Fukushima type accidents would all never have happened if we had MSR reactors instead for civilian usage. But since US Navy paid for all of the initial and follow on research on water cooled reactors, the private sector wasn't interested in investing billions on MSR or Sodium fast reactors. In the end, the reason the USA pursued Sodium fast reactors (aka fast breeders) was the perception the world was short on Uranium, but that has since been proven wrong (we have 100+ years of Uranium even using low burn-up reactors. But with Uranium or Thorium breeder reactors we move from having 100+ years of Uranium to having 10000-30000 years worth of Uranium.

[Robotbeat]

This wasn't the first time that the military conservativeness did us a sizeable diservice picking inferior technology.

Unsubstaniated

When you consider part of the reason that the military chose the water cooled reactors, not just for simplicity but to also partially process radioactive materials into materials bthat could be used in nuclear weapons, using a liquid metal reactor makes little sense.

But what the previous commenter failed to realize is that the vast majority of nuclear powerplants were to be used around water, the use of liquid sodium reactors is damned near suicidal!  Sodium in water is a bad thing as sodium explodes in water.
Moltent, liquid sodium, exposed to excessive amounts of sea water would cause an explosion and a potential nuclear accident of such a degree that Chernobyl would seem like a minor microwave malfunction.

I have to agree with Jim on this one.  Between nuclear incidents that have been declassified and ones that will likely remain classified forbthe rest of our lives, the choice that the military has made in the type of nuclear power system that the currently use, is the best and safest choice that could be made with the technology of the beginning of the nuclear age.

Why are you STILL talking about liquid metal, especially sodium??? Everyone here is talking about molten SALT. SALT, not sodium or any liquid metal.

[Nilof]

The fuel salt is temperature self-regulating, so uncontrolled salt will not be hotter than in routine operation.

Confused by this, seems incorrect. The fission reaction can have a temperature coefficient of reactivity, but not decay heat. That happens according to half lives of the isotopes, nothing can be done about it other than try to remove the heat by safe means before it removes itself by unsafe means.

The heat from short lived decay products is much lower than that produced from Fission, and molten salts conduct heat extremely well, so getting them hot is difficult and requires a compact geometry. Build the containment vessel out of materials with a high thermal conductivity and escaped fuel salt will tend to solidify.

Because of commonality with the coolant, you can also opt for active measures where you disperse the coolant into the containment vessel to dilute any fuel salt that has escaped. Or you could have containers filled with pellets of solid FLiBe ready to be spread out for the same end.

Either way, my argument was not that FP's don't heat up fuel. My argument is that FP's that can evaporate will do so in regular operation. So a fuel salt leak is not nearly as dangerous as a solid core meltdown because the actual problem cause by the latter, volatile FP's escaping and causing damage, cannot occur since they already escape during regular operation and don't build up. Remaining FP's that have a high boiling point are not nearly as much of a problem in the case of an accident.

Cooling is WAY harder in space because you're operating via radiative cooling instead of convective/conductive. Not so bad on Mars, but it's still harder than dumping heat into a river. In space, you need a big, heavy radiator, especially if you claim to operate at high efficiency.

But regardless, most space reactor designs use helium or something like that for coolant, not water. I don't know if ANY proposals I've read use water, in fact.

Indeed. Water cooled reactors are severely limited for space uses by their operating temperature. Radiator mass scales as one over T^4, so a lower temperature means that the radiator mass would get completely out of hand for a water cooled reactor. This is proably the main reason why nobody has been seriously considering them in space, with heavy high-pressure machinery likely being the other.

IMHO, the most feasible option is probably to focus on making something more like an RTG-replacement, like SAFE. Easiest to get permission for.

[macpacheco]

When you consider part of the reason that the military chose the water cooled reactors, not just for simplicity but to also partially process radioactive materials into materials bthat could be used in nuclear weapons, using a liquid metal reactor makes little sense.

But what the previous commenter failed to realize is that the vast majority of nuclear powerplants were to be used around water, the use of liquid sodium reactors is damned near suicidal!  Sodium in water is a bad thing as sodium explodes in water.
Moltent, liquid sodium, exposed to excessive amounts of sea water would cause an explosion and a potential nuclear accident of such a degree that Chernobyl would seem like a minor microwave malfunction.

I have to agree with Jim on this one.  Between nuclear incidents that have been declassified and ones that will likely remain classified forbthe rest of our lives, the choice that the military has made in the type of nuclear power system that the currently use, is the best and safest choice that could be made with the technology of the beginning of the nuclear age.

You fail to realize that most liquid metal (and even some proposed MSR) coolant reactors keep the fuel in solid form. So if the issue was being able to easily extract plutonium from the reactor core fuel, that would be just as easy as on water cooled and gas cooled reactors.

One reason to keep the fuel solid is to control exposure of the internal containment to fast neutrons, forcing the fast neutrons to pass through a moderator barrier before they can reach the inner containment walls.

[ArbitraryConstant]

My argument is that FP's that can evaporate will do so in regular operation. So a fuel salt leak is not nearly as dangerous as a solid core meltdown because the actual problem cause by the latter, volatile FP's escaping and causing damage, cannot occur since they already escape during regular operation and don't build up.

AFAIK Cesium is not removed continuously by evaporation, would require external fluid reprocessing. Nothing I've seen supports this. Do you have more information handy?

Cesium-135 is removed by virtue of its precursor Xenon-135 being removed, but in solid fuel Xenon-135 would quickly get converted to stable Xenon-136...

[Robotbeat]

How relevant is this to SpaceX, now?

[ArbitraryConstant]

Good point, I'm out.

[DarkenedOne]

A valid question is 'if not nuclear, then what?'

Outsource the nuclear part to mother nature; solar thermal propulsion

Thermal propulsion of any sort still basically needs liquid hydrogen to ever be better than chemical. And considering how much liquid hydrogen you need, probably greater delta-v could be had if you added the stoichiometric amount of oxygen and used it as a chemical rocket instead of a (solar/nuclear) thermal rocket.

ISP is the impulse per unit fuel.  The whole reason why nuclear propulsion is attractive is that it offers significantly higher ISP, then any chemical propulsion means available, therefore no you could replace a nuclear rocket with a chemical one, and expect to get higher ISP.

As far as needing liquid hydrogen to be better than chemical propulsion, chemical propulsion needs liquid hydrogen to achieve its highest performance as well.  Chemical and nuclear rockets need hydrogen for the same reason.  It is the lightest gas.   

As far as "outsourcing" nuclear power to the sun the problem is that the rate at which one could accumulate energy from the sun is significantly slower than one could get from a nuclear reactor.  Using solar power is ok for space probes in deep space that can hibernate for decades, but it is not ok for sending humans. 

[Robotbeat]

A valid question is 'if not nuclear, then what?'

Outsource the nuclear part to mother nature; solar thermal propulsion

Thermal propulsion of any sort still basically needs liquid hydrogen to ever be better than chemical. And considering how much liquid hydrogen you need, probably greater delta-v could be had if you added the stoichiometric amount of oxygen and used it as a chemical rocket instead of a (solar/nuclear) thermal rocket.

ISP is the impulse per unit fuel.  The whole reason why nuclear propulsion is attractive is that it offers significantly higher ISP, then any chemical propulsion means available, therefore no you could replace a nuclear rocket with a chemical one, and expect to get higher ISP.

actually both chemical and nuclear thermal rocket engines are significantly constrained by material properties and temperature limits. NTR needs to conduct heat through a solid heat exchanger to the gas, but chemical rockets can have the gas combust without contacting any solid surface directly, so for the same materials you can tolerate HIGHER temperatures with a chemical rocket. The big thing ISN'T the energy per kilogram of fuel but instead the molecular weight of the fuel in combination with energy. A typical hydrolox chemical rocket already tends to be constrained by temperature, so adding nuclear power would ONLY add more mass if operating at the same mixture ratio.

As far as "outsourcing" nuclear power to the sun the problem is that the rate at which one could accumulate energy from the sun is significantly slower than one could get from a nuclear reactor.  Using solar power is ok for space probes in deep space that can hibernate for decades, but it is not ok for sending humans.

Im assuming you're talking about solar electric compared to nuclear electric. If so, then you obviously haven't done the math. Look at any space reactor that has ever flown (Russian ones, especially), make sure to include reactor, turbine, radiator, shielding mass, etc, and compare it to current solar power and you see that solar power produces WAY more power per kilogram than nuclear, even to Mars orbit. At Earth orbit, the difference is like 10:1 in favor of solar power. It is ONLY in deep space far away from the Sun that this reverses, so in fact the very opposite of what you said.

FWIW, Ultra-Flex solar can do ~150W/kg. thin film solar on a lightweight substrate and structure can do >1000W/kg, like IKAROS showed. Compare with numbers for FLOWN reactors, not just feelings of what you think /sounds/ more powerful.

[MikeAtkinson]

Are there any MSR designs for space? If so have they gone through peer review from both space and nuclear engineers?

[macpacheco]

Are there any MSR designs for space? If so have they gone through peer review from both space and nuclear engineers?

AFAIK there are no finalized reactor designs for space use,just old R&D programs and preliminary studies, regardless of type of reactor. I believe there is not even a preliminary study for space MSR, since the last time NTR and nuclear space craft were seriously though of was 40+ years ago..
I tried to emphasize MSR reactors cause they offer the highest power / weight ratio, highest safety potential and they can be run with thorium which is much easier to find than uranium (thorium has a unique electromagnetic signature, using that signature Thorium reserves on the moon have been mapped a while ago), at the same time MSRs can be run on Uranium, Plutonium, Thorium and most mixtures you can think of, due to excellent neutron economy (critical factor in a reactor to be able to burn more of its nuclear fissile material and to convert fertile material into fissile).

[Robotbeat]

Jason: Actually, you're wrong. Thin film PV, in combo with either regen fuel cells or (more relevant nowadays with recent advances) state of the art Lithium Ion or Lithium sulfur (both of which are better than older regen fuel cells) beats nuclear power pound for pound and volume stowage wise for surface power on Mars. See this paper: http://systemarchitect.mit.edu/docs/cooper10.pdf

And in-space, PV trounces nuclear (ie how much power for a given mass) until you get past the asteroid belt. It's not even fair, solar is like 5-10x more powerful (if you compare existing or historical in-space nuclear to existing solar, OR credible new developments for nuclear compared with credible new developments for solar). That's why no one has nuclear powered satellites anymore.

The problem with PV on Mars is that your PV panels build up an electric charge from both use as well as dust storms, after a while, simply brushing off the panels doensn't work so well because the dust is now electrostatically stuck on the PV cell faces.  Plus, the fine dust would start to cloud the surface of the PV cells just from simple abrasion.  This is a small part of why the Mars rovers using PV cells are slowly but surely becoming unable to generate power.

    MIND YOU, this has not happened NEARLY as fast as anyone at NASA expected, thus the decade plus mission on a rover that was supposed to only last 90 days.  But there has, over the years, been a noticable and steady drop off of power that the cells can generate.

Maybe to some degree, but IIRC a cleaning event earlier this year got Opportunity up to 94% of its original capacity. So most of the loss does seem to be easily removable dust.

(And Opportunity apparently tends to get cleaning events on hillsides, so static panels placed on hills would probably do better than Opportunity has, since they would be on the hill all the time.)

Do you have a source about the 94% result? I'd be very interested in it.

[Vultur]

Jason: Actually, you're wrong. Thin film PV, in combo with either regen fuel cells or (more relevant nowadays with recent advances) state of the art Lithium Ion or Lithium sulfur (both of which are better than older regen fuel cells) beats nuclear power pound for pound and volume stowage wise for surface power on Mars. See this paper: http://systemarchitect.mit.edu/docs/cooper10.pdf

And in-space, PV trounces nuclear (ie how much power for a given mass) until you get past the asteroid belt. It's not even fair, solar is like 5-10x more powerful (if you compare existing or historical in-space nuclear to existing solar, OR credible new developments for nuclear compared with credible new developments for solar). That's why no one has nuclear powered satellites anymore.

The problem with PV on Mars is that your PV panels build up an electric charge from both use as well as dust storms, after a while, simply brushing off the panels doensn't work so well because the dust is now electrostatically stuck on the PV cell faces.  Plus, the fine dust would start to cloud the surface of the PV cells just from simple abrasion.  This is a small part of why the Mars rovers using PV cells are slowly but surely becoming unable to generate power.

    MIND YOU, this has not happened NEARLY as fast as anyone at NASA expected, thus the decade plus mission on a rover that was supposed to only last 90 days.  But there has, over the years, been a noticable and steady drop off of power that the cells can generate.

Maybe to some degree, but IIRC a cleaning event earlier this year got Opportunity up to 94% of its original capacity. So most of the loss does seem to be easily removable dust.

(And Opportunity apparently tends to get cleaning events on hillsides, so static panels placed on hills would probably do better than Opportunity has, since they would be on the hill all the time.)

Do you have a source about the 94% result? I'd be very interested in it.

"in May Opportunity's solar array dust factor went from 0.832 to 0.962, which is close to as good as it can get and a record for a rover more than 10 years into its mission."

http://www.planetary.org/explore/space-topics/space-missions/mer-updates/2014/05-mer-update-opportunity-hunts-ancient-clays.html (it's next to the picture of a hypothetical early Mars with oceans)

So it's actually 96%, even better.