Let’s say someone wanted to make a nuclear reactor to make electricity, what is the least demanding version of that? What would it look like?
More specifically, if one wanted to make a CANDU reactor, what is the smallest amount of natural uranium and heavy water that would be enough to produce electricity for, say, about 100 people?
I believe some deep space probes have small nuclear reactors.
Although you have to specify the purpose. A reactor suitable for spacecraft is not useful for naval vessals which are not ideal for utilities such as electricity production.
SNAP or RTG generator cores can be the size of waste baskets. They simply convert the heat from radioactive decay into electricity by putting bands of different metals around the core. Power is limited to a few hundred kilowatts.
Utilities for electricity production. I’m looking for the one which would be within easiest grasp of dude-in-a-garage.
Decay based “nuclear batteries” like that can be made far smaller, but aren’t technically nuclear reactors since they passively rely on the decay of their radioisotopes.
David Hahn made a nuclear reactor in his mother’s garden shed. This articlehas pretty detailed information on how he did it.
To Build? The problem is safely handling the reactive material in quantities that make power generation possible. I think that’s gonna be outside the ability of a dude-in-a-garage.
Perhaps with some kind of kit that came with a sealed core it could be feasible perhaps.
I’d like to make a correction: Relying on nuclear decay is fine. I would include that as well. As long as it’s some kind of radioactivity that would allow electricity production for a long time and be simple.
Safe handling not required. Presume the guy-in-a-garage has little to lose.
Would it be possible to put several dozen pounds of natural uranium in heavy water to get a subcritical reaction that would produce enough heat to generate electricity?
I believe they are more like a few hundred watts, not kilowatts.
The problem with trying to provide power for 100 people is that you just aren’t going to get any heat out of a core unless you are operating in the power range, critical, with fission in balance with neutron production. And for that many people you will need coolant, coolant pumps, and power generation equipment.
What you are describing is similar to the mission of the Army’s SL-1 Reactor. They wanted a small simple reactor that could be flown into remote locations (e.g. Antarctica or Greenland), set up, and maintained by minimal staff with minimal training.
The challenge with any reactor design is that power always changes at an exponential rate. Even when a gigawatt-size commercial plant has been shut down for weeks and they are bring it up from “cold iron” the tiny infinitesimal spark of fission going on increases at a exponential rate as they are starting up. One flashlight bulb of power… two… four… eight… until they reach the power. Much of this happens below the detection range of instruments, making a startup more interesting than it should be some times. This makes power excursions of any type very fast and troublesome.
Anyway, in the SL-1 reactor, they had just done maintenance on it, changing out some instrumentation, and three guys were hooking things back up to the reactor vessel. One of the guys was supposed to manually pull the single control rod up a quarter inch or so in order to make connections. Perhaps it was stuck, because he managed to yank it out several inches.
The tiny reactor had a power spike in the gigawatt range, resulting in a steam explosion and the deaths of the three workers.
It looks like the Army wasn’t able to get “simple” down to a science.
watts, not kilowatts… 
There was a small nuclear reactor in the basement (under the basement) of one of the buildings in the downtown Toronto campus of the University of Toronto. IIRC from the description, it was smaller than an oil drum and buried several feet underground.
It qualified as a reactor necause the mass of radioactive material was big enough to create a higher level or radioactive decay than in just small amounts of uranium. However, it was used for making isotopes and tracer samples, I think, not for power or anything. A capsule with he sample would be pushed down into the core, exposed for a timed duration, and retrieved. The material could be used for medical imaging, cancer treatment, etc.
What about the nuke powered car that IIRC Ford wanted to produce? I think it’s name was the ‘Nucleon’?
Tangential to the OP.
How big would a nuclear reactor be (including shielding) to meet the electrical needs for a house for a family of 5?
Well, the link isn’t about subcritical decay heat; this was a natural critical sustained fission reaction in a uranium ore deposit. It exceeds the expectations of your question.
Well, this can’t happen anymore.
The only way that this can occur naturally is for the U-235 concentration to be in the 3% range, which is far greater than it is now (due to the relatively short half-life of U-235).
Kodak also had some sort of small reactor in New York State.
Declan
Do I understand correctly that a subcritical uranium pile in heavy water will emit more heat than is typical?
Do I understand correctly that if one reduced the amount of water in spent fuel pools (and made it heavy water), the pool could get warm enough to boil?
This guy seems to think so: Using Nuclear Waste Heat as Power Source
Beowulf,
From what I understand, non-enriched uranium will not do much when you pile it up in light water. In heavy water though, natural 0.7% uranium will behave as if it were enriched.
Also, again, what’s the smallest amount of natural uranium in heavy water that could go critical? I know it’s unsafe and a bad idea. I’m not planning to do it.