Will SpaceX ever go nuclear propulsion?

[gospacex]

Radioisotope power supplies actually trade very well compared to fission from a specific power standpoint. Standard RPS/RTGs get 6-10W/kg compared to about 3-4 W/kg for the fission reactor on NASA's 5.0 Mars DRA.

This looks wrong. IIRC fuel is run at ~50 kW/kg power level in today's reactors.

[JasonAW3]

Too bad we can't take rock and dust from asteroids, melt it with Nuclear heat, and fire it out as an ion stream of matter.  Silicon, iron and nickle should give quite a boost with the use of a solar powered, continious stream linear accelerator.

[ncb1397]

Nuclear thermal rockets that don't use hydrogen have the low Isp of chemical rockets combined with the operational complexity and weight of a heavy rector and shielding.

Keep the reactor on the ground to produce propellant. Just because you COULD use a nuclear reactor doesn't make it a good idea. A lot of space enthusiasts treat nuclear power as space-magic, a wand to wave that makes everything magically feasible. If you actually look at the realistic engineering involved, nuclear thermal (and solar thermal for that matter) simply doesn't trade very well nowadays. In fact, I'd argue that the closer we get to routine, inexpensive space launch, the LESS sense nuclear thermal makes.

I don't see how this is a hard and fast rule. For instance, water has a density of 1 g/cm^3 while being 20% hydrogen or a hydrogen density of .2 g/cm^3. Liquid Hydrogen has a density of .071 g/cm^3 while being 100% hydogen or a hydrogen density of .071 g/cm^3. Water fed into a nuclear thermal rocket would simply go through thermolysis resulting in the constituent diatomic/atomic gases. Is there a reason why hydrogen that was derived from water gives you less impulse than hydrogen pure? It seems to me that storing as water would get rid of a lot of your density concerns. Not even counting the oxygen, water is more dense than liquid hydrogen while if for some odd reason you can't split the hydrogen in one step in the heating process, you could always use electrolysis to generate the h2 and then feed it into the NTR. In fact, this gets rid of the long term storage problem of using cryogenic fluids and would yield non-insulated tanks that are lighter than chemical schemes like LNG/LOX or LH2/LOX.

[Ludus]

By 2030 given the pace of R&D photovoltaics may be much more competitive. Not just output/weight but possibly suitability for local manufacture with ISRU for most of the mass.

If reactors are part of the mix ISRU is a consideration for them too. I think there is some data about Thorium availability.

For energy storage from Solar, some of the same issues apply as for nuclear produced electric. They will be making Methane from CO2 and water anyway for rocket fuel. They might also use new catalyst tech as from siluria to make liquid hydrocarbons both as easier to store fuels and feedstock for manufacturing.

If this is not just an outpost but the nucleus of a permanent civilization it will have to get down to making a wide range of stuff from local resources as soon as possible. One consideration is what energy tech is more flexible for home grown expansion.

[guckyfan]

We have threads for colony building in the Mars section. Please bring that discussion there.

On nuclear propulsion. I don't believe it will happen anytime soon by SpaceX or NASA. What the Russians will do, we cannot know. They are at least talking about it.

Ever or never are strong words though. Who knows what can be done in 30 to50 years?

[douglas100]

...Water fed into a nuclear thermal rocket would simply go through thermolysis resulting in the constituent diatomic/atomic gases. Is there a reason why hydrogen that was derived from water gives you less impulse than hydrogen pure...

Yes. The hydrogen that comes out of the nozzle is mixed with oxygen. The molecular weight of the exhaust gas mixture would be far higher than hydrogen on its own and results in an Isp much closer to the performance of a conventional LOX/LH2 engine. Add in the weight of the reactor and you lose any performance advantage over chemical propulsion.

[R7]

you could always use electrolysis to generate the h2 and then feed it into the NTR. In fact, this gets rid of the long term storage problem of using cryogenic fluids and would yield non-insulated tanks that are lighter than chemical schemes like LNG/LOX or LH2/LOX.

What would you do with the produced O2? If you just vent it overboard your effective Isp is 20% of what the hydrogen produces in the NTR because the oxygen mass still counts as propellant consumed. Using the oxygen in separate NTR doesn't help much either, it's Isp is in the low 200s and combined Isp isn't superb either. (=400s assuming optimistic 1000s for H2 and 250s for O2)

[JasonAW3]

By 2030 given the pace of R&D photovoltaics may be much more competitive. Not just output/weight but possibly suitability for local manufacture with ISRU for most of the mass.

If reactors are part of the mix ISRU is a consideration for them too. I think there is some data about Thorium availability.

For energy storage from Solar, some of the same issues apply as for nuclear produced electric. They will be making Methane from CO2 and water anyway for rocket fuel. They might also use new catalyst tech as from siluria to make liquid hydrocarbons both as easier to store fuels and feedstock for manufacturing.

If this is not just an outpost but the nucleus of a permanent civilization it will have to get down to making a wide range of stuff from local resources as soon as possible. One consideration is what energy tech is more flexible for home grown expansion.

Unless there is a serious change in how solar cells work, there is little chance that they will ever be on par, volume or mass wise tio Nuclear power.  Expense wise, that may be totally different.  but that's for another thread to discuss.

[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.

[simonbp]

...unless you're building it with a nuclear reactor in it. In that case, forget it! Won't happen. Will explode the cost and the weight. Good for Mars surface power, bad for Mars rockets.

Why bother with nuclear for Mars surface power? There is still plenty of sunlight on Mars, and if an astronaut-janitor can come around occasionally clean the dust off the panels, they should last quite a while. Plus, it is much more plausible that you could set up a low-efficiency solar cell production facility on Mars than building new reactors.

A single reactor might make sense as a pilot ISRU plant (since sabatier requires heat more than electrons), but after that, solar thermal collectors could take over. And if you're clever enough, you could do the pilot plant with solar thermal, and never have to deal with reactors at all.

I think nuclear power has a great future on Earth, but in space it's almost always more trouble than it's worth.

[guckyfan]

The sabatier reaction is exothermic. Only some initial heating is required to get it started. After that it produces more heat.

[sheltonjr]

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.edu

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.

Note:  Link to cooper10.edu is broken

I believe almost all past studies of nuclear power in space are using scaled down versions of solid fuel Light Water Reactors (LWR). As can be seen on this thread, Those that are for nuclear in space and on Mars are promoting the Molten Salt Reactor (MSR) as the nuclear technology that should be used.

LWR while good for terrestrial power generation have some characteristics that make them ill suited for space applications.

1) With water as the coolant and moderator, it must be under high pressure to obtain decent temperatures for power conversion. This requires a big and heavy pressure vessel.

2) LWR require initial excessive fission material and usable poisons and control rods to control the reactor. Limited fuel life and burn-up due to fission product build up. Extra backups and controls add weight and cost. Refueling is difficult or not available.

3) Water for heat transfer requires large/heavy heat exchangers

4) Relative low temperature of coolant requires large radiators for heat rejection.

MSR have the following characteristics that make them good for space applications.

1) Molten Salt has a higher heat transfer properties that enables small core, pumps, piping and heat exchanger components.

2) Molten Salts remain liquid past 1500^oC at atmosphere or lower pressures. No pressure vessel required.

3) The high operating temperature of 700-800^oC reduces heat exchanger and makes it possible to use a really small light weight super-critical CO2 turbine. (Needs to be developed)

4) The high temperature also greatly reduces the required radiator size required for waste heat rejection.

5) MSR have a negative coefficient of expansion that makes them self regulating. No excess fission inventory required in the reactor. Fission product gases can be vented and fuel can be slowly added as required to compensate for other fission products that cannot be removed easily.

6) High fuel burn up 70-90% requires less fizzle fuel. Long 30 year life span is possible. 


For under a 100KW with near continuous access to the sun,  < 24 hours, I think solar has the win.  But once we start talking multi-MW Electric Propulsion, ISRU and supporting a growing population on Mars/Asteroids/Moon MSR nuclear has some advantages that should not be ignored.

I just want to make sure to the forum that when we are comparing and speculating on solar and nuclear in space we are specific in which technology we are debating, and not making general statements that reference only technology that supports your arguments.  That would be like me basing all future solar power specification in space on what the ISS is using which we know is last generation technology.

[Robotbeat]

Fixed the link!

[Robotbeat]

Solar wins beyond 100kW, too. Thin film solar, especially, offers specific powers that nuclear can only dream of until you get well past Mars.

[guckyfan]

Can you throttle down or switch off a MSR?  I think you cannot have the molten salt go solid. That would make it very problematic for propulsion purposes I imagine.

[Nindalf]

I believe almost all past studies of nuclear power in space are using scaled down versions of solid fuel Light Water Reactors (LWR). As can be seen on this thread, Those that are for nuclear in space and on Mars are promoting the Molten Salt Reactor (MSR) as the nuclear technology that should be used.

I believe it's not good to get your information primarily from enthusiasts of a particular technical approach, who tend to gloss over all the facts in their eagerness to support their cause.

I'd call this a fairly typical example of a proposed nuclear reactor design for use in space:
https://en.wikipedia.org/wiki/SP-100
Hardly an LWR.  There's no drop of water in the system.  Solid fuel, solid moderator, liquid metal cooled, with thermionic power generation.

Or this:
https://en.wikipedia.org/wiki/Safe_Affordable_Fission_Engine
Solid fuel, solid moderator, liquid metal cooled, with a xenon-helium turbine for power.

Or this actual realized concept:
https://en.wikipedia.org/wiki/TOPAZ_nuclear_reactor
Solid fuel, solid moderator, liquid metal cooled, with thermionic power generation.

Molten salt vs. liquid metal is certainly an interesting choice to study (although they went head-to-head at one point, and liquid metal won, and has been accumulating engineering experience through routine use ever since), but it's hardly a decision likely to result in dramatic improvements.

If you're looking for revolutionary improvements in power-to-weight, you've got to look at really exotic designs like dusty plasma fission fragment reactors, where finely divided fuel serves as the main radiator, and the power output is in the form of relativistic charged particles.

[sheltonjr]

Can you throttle down or switch off a MSR?  I think you cannot have the molten salt go solid. That would make it very problematic for propulsion purposes I imagine.

I admit I am not an expert but have been investigating it for a few years.

The way I understand it is that the fuel salt has a negative temperature coefficient of expansion, so when the salt heats up it expands like most liquids. This expansion is enough to reduce the quantity of fissions in the fuel-salt to reach a stead-state temperature based on the quantity/density of fizzle material in the fuel-salt.

Therefore the reactor is self regulating or more accurately Load Following.  Need more energy, Run the coolant/power loop faster which cools the fuel salt and increases fission. Need less power, Slow down the pump. It will never solidify at min power because there is always enough fissioning happening to maintain the equilibrium temperature.

I do not see why not this will not work down to 0% power.  The system will have to be designed so that the steady state temperature with the designed fizzle inventory does not exceed the container max temperature specifications.

Once every couple weeks/months a M&M sized fuel pellet will be added to the reactor to make up for used fizzle material and fission product neutron poison (absorbers).  A 1.5MWe reactor uses about 1.5 grams of uranium per day at full power. 1 KG will last 2 years.

[RanulfC]

you could always use electrolysis to generate the h2 and then feed it into the NTR. In fact, this gets rid of the long term storage problem of using cryogenic fluids and would yield non-insulated tanks that are lighter than chemical schemes like LNG/LOX or LH2/LOX.

What would you do with the produced O2? If you just vent it overboard your effective Isp is 20% of what the hydrogen produces in the NTR because the oxygen mass still counts as propellant consumed. Using the oxygen in separate NTR doesn't help much either, it's Isp is in the low 200s and combined Isp isn't superb either. (=400s assuming optimistic 1000s for H2 and 250s for O2)

The O2 would be added to the hot hydrogen exhaust for thrust boost most likely and avoid putting it through an NTR at all. LANTR (LOx-Augmented NTR) concepts tend to have lower ISPs than pure LH2 propellant but the extra T/W is nice to have.

Thing is the whole electrolysis concept is probably too power intensive to be practical. You'd use less power going with a tri-mode NTR (Propulsion, Augmented-Propulsion, Power Generation) NTR with cryo-cooling capability rather than carrying the electrolysis gear. IIRC however concepts like the "Kuck Mosquitoes" or similar use a similar concept to move water/ice around the solar system pretty efficently.
http://www.spacefuture.com/archive/the_deimos_water_company.shtml
http://www.projectrho.com/public_html/rocket/infrastructure.php

Randy

[ArbitraryConstant]

I believe almost all past studies of nuclear power in space are using scaled down versions of solid fuel Light Water Reactors (LWR).

Incorrect.

Reactors proposed for use in space, not to mention actually used in space, are predominantly (all?) either metal or gas cooled. Also predominantly (all?) HEU fueled. They're really not that close to any widely used terrestrial reactors. Probably the closest would be eg the metal cooled fast reactor used on some high performance Soviet submarines.

[go4mars]

I know this is quite speculative, but following reading 'The Martian' by Andy Weir, and SpaceX's push for Mars, got wondering, is there any merit to go from chemical rockets to nuclear at some point in the future?

Well, I will suggest that the main reason for Jeff Bezos investment into the company General Fusion, isn't earth-based fusion power.  Rather, it's follow-on iterations for extreme isp in-space propulsion.  My speculation. 

Is Elon thinking along these lines too?  Not demonstrably that I've seen.