Hacker News Reader

Comment by Robotbeat | Hacker News Reader

The headline is potentially misleading. There are already "nuclear power plants" based on radioisotope decay of Pu238 on Mars (and old ones on the Moon) as well as beyond, such as Voyager 1, 2, and New Horizons (all of which are still operating). They're called RTGs.

In order to get approval to launch nuclear material, the launch vehicle must be nuclear-rated (like human-rating, it ensures safety). With the recent retirement of the Delta II, only the Atlas V is currently nuclear-rated, but eventually others may be. Additionally, the fuel is encased in boxes designed to withstand launch failure and impact into the ocean without releasing the contents.

In the case of RTGs, the fuel is always releasing heat and is always radioactive. For fission reactors, the fuel is relatively inert and non-radioactive until it is turned on.

View on HN

Actually the RTG fuel capsule thermal and mechanical protection is so good, that when the Nimbus B satellite fell into the ocean due to a launch failure, they recovered it's RTG power source from the ocean floor and relaunched it on Nimbus 3: https://en.wikipedia.org/wiki/Nimbus_B

If I recall correctly, the RTG that was on Apollo 13's Lunar Module also survived reentry (since they dumped it into Earth's atmosphere after using it as a life boat). It's intact somewhere on the bottom of the ocean.

Tangentially,

CIA and IB (Indian Intelligence) lost such Plutonium RTG powered listening device at a Himalayan peak in 1965 used for SIGINT against China. Further expeditions to retrieve the device resulted in failure and later they installed another such device in 1967[1] which was supposedly taken out in 1968.

[1]https://www.livemint.com/Leisure/3QfYqLadggrbnrn41H0mAJ/The-...

I remember reading about that in the Wikileaks disclosures, it's fascinating to think about.

It was publicly disclosed at Indian parliament by PM in 90's after a press leak I think, not sure what Wikileaks disclosed about it.

I think they are going for something more than a simple RTG. I read a few years ago about Stirling engines in zero/low gravity. I think that is what they are going for, something with moving parts, rather than a relatively inefficient static RTG.

Yeah, it’s called Kilopower. A small fission reactor using Stirling engines. Up to 10 kilowatts electric

Can I have two for my house?

Half a century ago some companies actually planned for things like that - a small nuclear reactor that you would bury under your house and it would provide you with power for decades. It went the same way the nuclear car did - the concept would work but it's just too dangerous without constant supervision.

I had never heard of that, but the idea of it makes me think of all my neighbors growing up in the 90s who were spending tens of thousands of dollars cleaning up contamination from leaking underground heating oil tanks. It was a major issue for people buying and selling older houses...sellers hiding knowledge of leaking tanks, surprise discoveries etc. In the worst cases the soil decontamination could cost over $100k. I can only imagine what fun these end-of-life home nuclear reactors would be causing 50 years later!

I visited a home once that was built in the mid 20th century on the assumption that electricity would become too cheap to meter. The original heating system was described to me as just a bunch of resistive elements strung in the attic. I think the house actually got an award similar to LEED (or maybe it was LEED?) at its construction.

My parent's late 70s house has something like this. There are heating elements sandwiched between two layers of drywall in the ceiling. It actually works really well, because it heats objects rather than the air--sort of like the sun on a cool, cloudless day--but electricity is like 4x the cost of natural gas so it never gets used. Also, the attic has insufficient insulation, so half the heat would go the wrong way.

Oh yeah my (boarding) high school had a dorm built in the “electricity will soon be too cheap to meter” era and had heating elements set into the concrete floors. It housed like 50 students and used more electricity than all of the other buildings on the 500 student campus combined. The warm floors were very nice though!

What on earth do you need 20 kilowatts continuous for? That's a huge amount of power for domestic use.

Typical US domestic service is 24kW (120V/200A nominal at the main breaker).

Fair enough. Still, it's one thing to be specced for 24kW peak load (the breaker) and quite another to require 20kW continuous (two 10kW nuclear power plants). I still think 20kW is an awfully huge amount of power - the average US house only uses about 1 kilowatt when you smooth it out.

What on earth do you need more than 640k of memory for? That's a huge amount of memory for domestic use.

Nobody knows what kind of innovations a huge readily available sustainable energy supply would allow, but I bet they'd be cool.

(1) Heat. Space is cold. (-100c)

(2) Cooling. Space is hot. (+400c)

(3) Running the computer.

(4) Running the communications gear.

(5) Cooling 4+5.

How can you use electricity to cool things in space?

Piezo-electric? But then you need to cool the other side even more..

You run a pump on a coolant loop, just like any air conditioner.

Same as in a car or any air conditioner. You pump a hot fluid to/from radiators. Look at the iss. See the big black structures always held perpendicular to the solar panels. Those are the radiators.

https://en.wikipedia.org/wiki/External_Active_Thermal_Contro...

You build your walls out of Peltier elements. On one side of capsule, hot side will be in, cold side out, the other side of the capsule, the reverse. And then you use the generated power to drive a laser to shoot the energy right back into the sun.

I'd be amazed if GP's house is in space. But hey, life's full of surprises!

I'm less familiar with Kilopower but I spent a couple years working on something very similar that ran off of PU-238. Looks like Kilopower is U-235.

Looking into this, it needs a simple control system. The PU-238 bricks just made heat continuously with no control. But, they were also much less power.

If there's a control system you can launch the reactor cold. With the Plutonium bricks you cannot, meaning your cooling system (engine in this case) has to keep running during launch.

https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/200900...

https://rps.nasa.gov/technology/

An rtg can be designed to handle cooling failure. It isn't like a water reactor. As the brick heats it expands, decreasing reactivity and heat production. So an rtg can live perfectly well without its cooling system. It won't be making useful power, but it isn't going into meltdown either.

The RTG has to be kept cool continuously on the launch pad with a nitrogen purge. All cooped up in a fairing, it’d cook the spacecraft if the heat wasn’t rejected somehow

This wasn't a typical RTG. It had stirling engine and was a lot more efficient (made a lot more power per unit weight of Pu) than RTG.

I don't think there was any way to save and restart anything if an engine failed as it would have likely destroyed the engine.

https://rps.nasa.gov/system/downloadable_items/36_APP_ASRG_F...

That would be nice down in the tunnels on Mars. The waste heat would be used to heat the habitat.

The Russians used RTGs in various Siberian lighthouses too[1]. Unfortunately after the fall of the USSR they were looted, RTGs and all.

[1] https://englishrussia.com/2009/01/06/abandoned-russian-polar...

This is exactly the issue I see with nuclear. We're creating dangerous waste products that last tens of thousands of years (normal nuclear - I know RTGs don't stay radioactive nearly as long!), and are easily weaponised.

Yet on the other hand we can't even keep a country stable for more than a few decades. The soviet union fell into disarray that time leading to all kinds of incidents, and we had a major world war only 75 years ago which is nothing on nuclear waste time scales.

One way or another this will bite humanity in the back in the future. If it isn't by a malicious future actor weaponising it, it'll be by neglected nuclear waste storage facilities that last much shorter than the waste itself.

For another example of things that can happen see https://en.wikipedia.org/wiki/Goi%C3%A2nia_accident . That one was medical in origin but the same can happen to nuclear waste when not adequately protected of which there is a lot more.

>The headline is potentially misleading. There are already "nuclear power plants"

It is not potentially misleading. This is just you being pedantic about a connection that very few people make. When you say "nuclear power plant", people tend to think of nuclear fission reactors, not RTGs or fusion reactors. Both of these are applicable to the term "nuclear power plant", but only one is commonly referred to by the term. A lot more people would feel misled if it actually referred to RTGs when it said "nuclear power plant."

The only group that it could potentially mislead is people who know RTGs exist. Most of those people would instead think "I wonder what kind of nuclear power plant it is" instead of "wow, what a misleading headline!" unless they were intending to be pedantic (like you).

I imagine* an RTG has a well enough encased piece of radioactive metal that will survive falling through the atmosphere without disintegrating.

Which makes the main danger that the Plutonium may get into the wrong hands.

* Happy to learn if that's wrong :)

The plutonium in an RTG is the wrong type from a weapons perspective.

The casing on RTGs have typically been designed to survive reentry. One example is Nimbus B - the rocket failed late in flight. The RTG was recovered from the ocean, refurbished, and flown again on Nimbus 3.

> The plutonium in an RTG is the wrong type from a weapons perspective.

You could potentially make a dirty bomb out of one. Just a conventional warhead that sprays radioactive fallout.

> Happy to learn if that's wrong :)

I’m not sure I’d be happy if you were wrong about this...

Reply on news.ycombinator.com