Short summary: Soviet engineers installed RTG powered radio relays to support the construction of a damn in Georgia. Political instability lead to the abandonment of the RTGs. Someone scavenged the generators and removed the radioactive cores from them.
Two of the radioactive sources were discovered by men gathering firewood in the forest. They decided to bring them to their camp(!) and cozy up to them to keep warm during the night(!!). Despite showing symptoms of radiation poisoning they kept the cores on their person while loading their truck(!!!). They all suffered terrible radiation injuries.
There are more sources "lost" from the same batch which remains unaccounted for to this day.
https://m.imdb.com/title/tt1588875/
Only the Soviets were daft enough to build RTGs using Strontium 90.
The United States did too.
Oak Ridge National Laboratory technical report "Strontium-90 Heat Sources"
https://technicalreports.ornl.gov/1971/3445605716035.pdf
Introduction
Compact electrical generators powered by heat from radioisotopes have been under development in the United States since the early 1950s for space, marine, and terrestrial uses. Essentially all the generators developed for marine and terrestrial uses have been powered by 90 Sr. This report summarizes the development work done by Oak Ridge National Laboratory (ORNL), Hanford Atomic Products Operation, and Martin Company, Nuclear Division, which led to the production of 90 Sr heat sources for use in the generators.
It was a natural choice since strontium 90 is an inevitable byproduct of operating any fission reactor, and was readily available as a coproduct from weapons plutonium production reactors. Making better RTG isotopes like plutonium 238 required additional infrastructure.
I mean, Sr90 is super cheap, and as long as it stays inside the RTG you're fine. The AEC actually tried Polonium RTGs in the late 50s.
The shorter half-life of Sr compared to Pu also means it's a bit less of an issue when you lose the source.
https://www.emergencylights.net/collections/self-luminous?gc...
They aren’t generating electricity though.
These batteries have very poor power density and are very inefficient. The advantages of nuclear-powered batteries are:
- They generate power over a long time, decades,
- They generate some heat.
They don’t generate much power. If you have a building, you would definitely think of a nuclear RTG as a “very shitty battery”, and that’s even if you don’t care at all about radioactivity.
Thinking of these as a “battery” is also a bit misleading, IMO. These are really just small power plants, which generate heat and turn the heat into electricity. The heat is powered by radioactive decay of Pu-238, and then turned into electricity with the extremely inefficient Seebeck effect. If you had a source of heat you wanted to turn into electricity, it’s much more efficient to use that heat to turn a turbine which is connected to a generator. And if you want an efficient, cost-effictive turbine, you make it big. At that point, you have a power plant.
The beta decaying substance is connected electrically to one electrode of a capacitor, while the electrons emitted due to the beta decay are able to pass through the insulating layer of the capacitor, reaching the other electrode.
Thus the capacitor is charged directly by the beta-decay and it can provide electrical energy to the external circuit.
The problem is that a nuclear reactor is a dynamic system, with some moving parts. It requires thermal management. It requires dynamic control. It is really hard to design a fully self-contained nuclear power system which wouldn’t require any human intervention to operate.
And even if we could, there is also a problem of waste management. Nuclear waste is not too dangerous, if you don’t touch it. It is, however, quite dangerous, if you grind it into fine particles and spray a large city with it by a crop duster. Our world is crazy. There are people like that out there, who might be interested in it. It is relatively hard to obtain hot nuclear waste from centralized large power plants. It will be really easy in the case of small building-scale reactors.
We'd need to have a lot of money, a disregard for return of investment and a lot patience: Current RTGs can do that, but they're rather expensive for heating houses and problematic from the nuclear materials POV (waste / profileration), not to mention the regulatory and licensing for using it a neighbourhood - better budget the time and money for lobbying for some legislation changes.
If by building we mean say 10 apartments, and each needs 10 kW, the RTG would need hundreds of kg of Pu-238 plutonim dioxide [1].
It's hard to cite the exact cost for that since it's not a freely traded commodity but that's a lot of plutonium. Eg NASA said that with a $75-90 million investment they can make 1.5-2 kg per year of it. [2]
[1] https://drinksavvyinc.com/blog/how-much-does-a-radioisotope-... gives 2 kW per 5 kg [2] https://www.space.com/20774-plutonium-spacecraft-fuel-nasa-b...
The EneFarms used to be heavily subsidized by the japanese government in a long term program to encourage fuel cell development and manufacturing. Over time prices have decreased such that the subsidy is either already expired or could be soon expired.
The tech is near, and allows getting a bit more energy out if natural gas. The gas companies hope it will allow them to eventually reuse their pipes to send hydrogen. Personally I think the combo of cheap solar panels and 400% efficiency heat pumps will outcompete gas.
If heat pumps are 400% efficient then log burners in cabins in the woods are even better.
Thank you for helping me learn a new HVAC term!
It's desirable for multiple reasons, of course, but it's not efficiency.
The 400% metric let's you compare with other heaters, the 100% is kind of useless.
A heat pump takes heat from outside the system. You put in x fuel and you get >x heat. Getting more energy than you put in makes the efficiency over 100%.
Even manufacturers call this 'coefficient of performance', not efficiency.
> "You can't [...]"
He did, so obviously he can. You mean shouldn't, not can't.
If we're willing to be so blasé with 'efficiency' then why not, say, 'functional programming'? If it works it's functional right?
I think only matter-antimatter reaction comes close to 100%. Burning fuels isn't even 1% of that.
Small research reactors exist, but they tend to generate in the neighborhood of tens of watts.
Only really viable in deployments that need very little power, where no other energy harvesting method is available, and periodically changing out batteries is not an option.