Will SpaceX use NASA Kilopower Nuclear Reactors?

[Robotbeat]

77% mass hit? To the whole power unit, which is already heavy? Or just the core?

[Space Junkie]

This is all very exciting. I'm actually just as interested in the technology for unmanned outer Solar System missions as for manned Mars missions.

I have one big question, though. Even if all the tests go well, what chance does this technology has of being funded through to a flight model? I hate to be the pessimist in the room, but this strikes me as one of those programs that gets cut as soon as there's a budget squeeze - particularly since there's no immediate need for this system.

Am I off base on this? Does NASA seem committed enough to this technology to see it all the way through?

(Maybe I'm just bitter about the ASRG program being defunded.)

[john smith 19]

77% mass hit? To the whole power unit, which is already heavy? Or just the core?

Assuming the LEU core duplicates the energy O/P of the HEU that would be the core only, including it's associated shielding. The Balance of Plant would remain the same, since it's producing the same output at the same temperature.

[BeyondNERVA]

I have one big question, though. Even if all the tests go well, what chance does this technology has of being funded through to a flight model? I hate to be the pessimist in the room, but this strikes me as one of those programs that gets cut as soon as there's a budget squeeze - particularly since there's no immediate need for this system.

Am I off base on this? Does NASA seem committed enough to this technology to see it all the way through?

NASA has always been interested in nuclear power, since they have always understood the inherent advantages. However, due to the way the government is structured, it's not just NASA's show, but the DOE's as well - and the DOE make NASA seem like daredevil risktakers. The DOE has also always been interested in in-space nuclear use, but hasn't had much funding for it. The director of Oak Ridge puts it like this: "We don't get a check that says Oak Ridge National Labs, $1.8 billion (probably the wrong number), go do good things." Everything's allocated by program, except for a small slush fund.

That small slush fund was just enough to get DUFF to happen, and now KRUSTY as well (less than $100 million, IIRC, it's in the post). These are the expensive bits. Now that the fuel fabrication has been proven out to Y12's satisfaction, and the reactor fueling and spacecraft integration have already been dress-rehearsed by NASA, the DOE, and the NNSA, we know how much it'll cost, to a much higher degree than many other new systems. That means that it's far more likely to fly, especially since it truly is a game changing development.

If we have a new flagship probe like Cassini, then this will probably fly. If we're doing Martian or lunar ISRU, this will quite probably fly. If we have Mariner III, or Cassini II, or whatever you want to call it, this will probably fly. Just like any other space component, this provides a capability with a set of engineering and financial tradeoffs. For the right mission, it's perfect, and a steal. For others, it's a cool piece of metal hanging out on the end of your spacecraft that you don't really need.

(Maybe I'm just bitter about the ASRG program being defunded.)

Yeah, I like the ASRG as well, but if you look at the recent GAO report (https://www.gao.gov/products/GAO-18-161T) about 238Pu production, you can see the problems that NASA's running into for RTG/ASRG fuel. The other advantage is that a reactor is radiochemically inert during launch, so assuming you can prevent accidental criticality during a launch failure (which they've done) you have no radiological concerns. This isn't necessarily the case with an RTG.

The ASRG's Stirling is actually running on KRUSTY. It still lives, even if as something like one of Terry Pratchett's Igors...

[Robotbeat]

77% mass hit? To the whole power unit, which is already heavy? Or just the core?

Assuming the LEU core duplicates the energy O/P of the HEU that would be the core only, including it's associated shielding. The Balance of Plant would remain the same, since it's producing the same output at the same temperature.

Right, but the 77% could already be (and quite likely is) taking that into account. That's why I want BeyondNERVA to respond.

[BeyondNERVA]

77% mass hit? To the whole power unit, which is already heavy? Or just the core?

That's the whole power unit. Because of the different distribution of energy within the reactor, and the different reactor geometry, there are different shielding requirements. The PCS, neutron shield, and heat rejection system stay the same, but everything else will need a little bit of a tweak. Unsurprisingly, fuel mass accounts for a lot of it, but gamma radiation shielding also accounts for quite a bit of it as well. There's a breakdown in the paper (https://fas.org/nuke/space/leu-reactor.pdf).

A 10 kWe unit for Mars ISRU, using the same fuel form, but LEU instead of HEU, weighs in at 2187 kg, instead of 1519. The equivalent space system would be 1945 kg vs 1120 kg. The 1 kWe units face a bigger hit.

All of the options (except advanced UZrH fuel-moderator composite) examined are metal fuel, which isn't ideal for LEU. It's just what's available for cheapest at the moment from NASA's suppliers. CERMET fuel may be a good option in the future, but a different fuel type means requalifying the fuel element... which is a pain in the butt. Nevertheless, you do have other options (at the higher power levels, oxides become the preferred fuel type, see toward the end of my post) for LEU fuel that may be easier on the mass budget!

This is very much still a first step, although an awesome one, even when it flies.

[speedevil]

I idly wonder if the licensing cost of launching a reactor is more expensive than stealing one of the many deactivated ones in orbit.
(above article goes into history, and is recommended.)

_{I suspect this is probably a silly idea.}

[BeyondNERVA]

I idly wonder if the licensing cost of launching a reactor is more expensive than stealing one of the many deactivated ones in orbit.

_{I suspect this is probably a silly idea.}

There were three types of reactors launched, SNAP 10A (which had an electrical bus failure), TOPAZ (used on the Soviet US-A RORSATs until 1990), and the BES-5 (launched twice in the late 90's-00's, and certified by NASA and the DOE). The American one is probably cold-welded to oblivion. The TOPAZ and BES-5 were both sodium-cooled, and were designed to eject the core, which all but three did (http://www.svengrahn.pp.se/trackind/RORSAT/RORSAT.html), so there's no way to start them up again.

There are two Soviet-built BES-5's at Los Alamos, IIRC. Originally, one of NASA's DRM's for a lunar base specified using one or both of those reactors for power. If we really wanted to, we could probably refurbish them and launch them... but I'm betting it would cost as much or more to do than just going ahead with Kilopower.

Here's some background on the TOPAZ International program: https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20140016877.pdf

[Patchouli]

This is all very exciting. I'm actually just as interested in the technology for unmanned outer Solar System missions as for manned Mars missions.

I have one big question, though. Even if all the tests go well, what chance does this technology has of being funded through to a flight model? I hate to be the pessimist in the room, but this strikes me as one of those programs that gets cut as soon as there's a budget squeeze - particularly since there's no immediate need for this system.

Am I off base on this? Does NASA seem committed enough to this technology to see it all the way through?

(Maybe I'm just bitter about the ASRG program being defunded.)

Even 10KW would  be a game changer for larger missions as you'd be going from half a hair dryer's worth of power to something comparable to entire house.

Data rates can be many times higher, a probe can have more powerful scientific instruments and much more powerful computers for autonomous operation etc.

[A_M_Swallow]

I talked with a SpaceX representative a few weeks ago about this given that I am a nuclear engineer.

They have essentially no realistic concept of how to refuel on Mars. {snip}

IMO with Kilopower on Mars you do not bother to refuel just shut it down and buy a new reactor.

With a submarine the old reactor core needs removing from the submarine to make room for the new reactor and core. With surface Kilopower just leave the assembly where it is. Install a replacement cable to get the power into the base.

10 year old Sterling convertors may be nearing the end of their life so replacing the whole assembly is the easiest way of replacing them.

[Robotbeat]

I wont tell you who I talked with. He was a propulsion engineer.

I read somewhere that Tesla, with Elon as CEO is fighting against subsidies needed for nuclear to compensate for the artificially low whole sale prices. Of course, Solar City and Tesla Powerwall wants all the subsidies in the world.

Dont get me wrong, IMO he is acting super anti nuclear. I cant find a reference sorry, I read it a while ago somewhere.

This is a very non-scientific way to approach the world. Do your research, validate your sources and don't propagate rumours.

Tesla apparently supported the shutdown of Diablo Canyon in California. I haven't independently confirmed it, but it made me mad (it's possible, but not very likely IMHO, that this was done without Musk). Musk at other times has been somewhat pro-nuclear. So saying he's "super anti-nuclear" is not accurate, but supporting the shutdown of Diablo is not very encouraging.

[Robotbeat]

Tesla (new user, not the company): You don't refuel space reactors. You just fill them up with all the fuel they'll need in their lifetime.

[Robotbeat]

I want to point out that sometimes the pro-nuclear community considers anything that could replace nuclear (such as storage-backed solar, if you oversize it a lot) as being anti-nuclear, which is dumb. Just because Tesla is enabling an alternative to nuclear doesn't mean Musk is anti-nuclear. I support both nuclear and renewables, and I realize that it is usually possible to do without nuclear (although northern nations benefit from nuclear a LOT, since seasonal solar output is anti-correlated with energy requirements for heat...).

...a mix is best.

[BeyondNERVA]

You don't refuel space reactors. You just fill them up with all the fuel they'll need in their lifetime.

That's certainly the way we do it now, but how is that different from every other piece of space hardware we've launched other that Hubble and the ISS? One benefit to the Kilopower design that doesn't get talked about (because of its' mission, no doubt) is that having a single fuel element, that's roughly cylindrical, could make refueling a snap. Isolating the used fuel element is a different problem, although due to the low power level, long mission life, and small amount of fissile material it shouldn't be TOO hot (but spent nuclear fuel is the most complex mess of isotopes known to humanity, and what comes out depends on a TON of factors that I don't even pretend to begin to understand).

Satellite refurbishment is already getting more attention. The problem is that we don't have a space tug to do the mission. Obviously, I'm a fan of a nuclear thermal-based infrastructure, but assuming lunar ISRU then ULA's ACES becomes very attractive. The other concept that I've always loved (which would probably give mission controllers headaches) is an old Rockwell design, the Solar Moth. Not super powerful, but you don't REALLY need that for refurbishing/refueling missions. http://www.projectrho.com/public_html/rocket/enginelist.php#solarmoth

Once it becomes possible, I'm pretty certain that swapping out fuel in space reactors WILL be a thing. They'll be very different designs, but as an example: current NTR designs aren't really set up for in-flight repairs, much less cracking open the core. This is something that would undoubtedly change once cislunar space becomes developed and manned interplanetary missions become infrequent news rather than the stuff of speculation.

The information I was looking for on the infrastructure needed for refueling reactors is buried in here (somewhere), I just can't seem to find it at the moment: http://www.projectrho.com/public_html/rocket/infrastructure.php

[Nomadd]

 Considering the logistics of refueling a core in space, and the fact that the core will be optimized size and mass wise, I'm not sure it wouldn't be much easier in every way to simply swap out the entire core every 30 years.

[BeyondNERVA]

Considering the logistics of refueling a core in space, and the fact that the core will be optimized size and mass wise, I'm not sure it wouldn't be much easier in every way to simply swap out the entire core every 30 years.

Depends on the reactor! If it's something like Kilopower, definitely, but Westinghouse Astronuclear had a design for what they called a PAX reactor, based on the NERVA A6 (about 2000 MWt, IIRC), that was designed to swap the core out fairly easily. There were other issues that would have made servicing a bit more challenging (graphite wool everywhere, for one), but nothing fundamentally unfixable. That doesn't seem to be the case with nuclear electric systems to nearly the same degree that I've seen, though - there's just never REALLY been a point. An in-core thermionic setup with a "flashlight" configuration should be fairly easy to refuel, though...

[acsawdey]

Considering the logistics of refueling a core in space, and the fact that the core will be optimized size and mass wise, I'm not sure it wouldn't be much easier in every way to simply swap out the entire core every 30 years.

Depends on the reactor! If it's something like Kilopower, definitely, but Westinghouse Astronuclear had a design for what they called a PAX reactor, based on the NERVA A6 (about 2000 MWt, IIRC), that was designed to swap the core out fairly easily. There were other issues that would have made servicing a bit more challenging (graphite wool everywhere, for one), but nothing fundamentally unfixable. That doesn't seem to be the case with nuclear electric systems to nearly the same degree that I've seen, though - there's just never REALLY been a point. An in-core thermionic setup with a "flashlight" configuration should be fairly easy to refuel, though...

Wasn't one of the design goals of Kilopower that it be easy and quick to fuel it (i.e. insert the fuel element) right before launch? Doesn't that imply that it might also be easy to open it up later and swap the fuel element for a new one?

[BeyondNERVA]

Wasn't one of the design goals of Kilopower that it be easy and quick to fuel it (i.e. insert the fuel element) right before launch? Doesn't that imply that it might also be easy to open it up later and swap the fuel element for a new one?

Possibly, although with the combination of the thermal load from the core and the thermal environment of space, I could see issues developing that would prevent removal.

You're correct about the fast fueling, however. It sounds like it needs less than a week from the reactor showing up on site (probably the VAB) to the shroud being able to be installed. At that point, it's really up to KSC how long they want it to sit around with all those guards...

[biosehnsucht]

The future documentary "The Fifth Element" covered refueling of nuclear-powered spacecraft :

Joking aside, it seems like not a hard problem to solve, if it was a problem worth solving. But I don't think it will be in the short term - and likely by the time it makes sense to solve it, other solutions will have occurred (at least for power on Mars or similar - build bigger reactors, or replace with some magical fusion, or whatever)

Though I don't think you'd leave a Kilopower module laying in place once done with it, you'd truck it out to be retired someplace out of the way - not because of radiation concerns from the old reactor, but because it's taking up valuable real estate and you could either install another one right there or expand the base / settlement a bit into where it was.

[BeyondNERVA]

Though I don't think you'd leave a Kilopower module laying in place once done with it, you'd truck it out to be retired someplace out of the way - not because of radiation concerns from the old reactor, but because it's taking up valuable real estate and you could either install another one right there or expand the base / settlement a bit into where it was.

It's quite possible that the real estate would be important, but the unit isn't that big, even with the surface radiators (which are approx. the same size as the space-based ones), and it seems like a lot of the designs call for a separation from the base greater than what NASA shows in its' animations. This is due to the dreaded ALARA, or "as low as reasonably acceptable," principle, which is fishy enough to pin down on Earth, but in space (esp. with NASA's current crew career dose limits) becomes a whole new slippery swimming pool full of hagfish.

If you're interested, much of the shielding and emplacement planning for Kilopower is based on an earlier design, the Affordable Fission Surface Power System (AFSPS, also sometimes called FSP), which Poston and McClure (the two nuclear engineers with the DOE spearheading this effort) both worked on. Some of the information is in here. I'm currently trying to track down their paper from NETS 2009 about this, but it doesn't seem to be published on either NASA or the DOE's websites...

http://www.neofuel.com/Poston-2009-NASs%20Lunar%20Fission%20Surface%20Power%20System-LA-UR09-02470.pdf

http://aip.scitation.org/doi/abs/10.1063/1.3115554

https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20110007114.pdf