Small Nuclear Reactor

[230RN]

My understanding is they basically produce power with no moving parts.  You have the loving soft warm glow of radioactive decay, which you surround with TECs (Thermoelectric Couplers).  TECs generate electricity as heat is transferred from one side to the other, but are not very efficient at all.  I played with TECs a few years ago when I experimented with building my own custom dimension fermentation cooler, and used TECs to cool refrigerant and expelled the waste heat through radiators and fans.  Used a lot of electricity in order to inefficiently move a little bit of heat.

The space implementations just have large radiators on the outside of the TECs rather than liquid cooled passages and pumps like my home very non-space-worthy application.  The more heat saturated a TEC gets, the less efficiently it operates.  In a thin or nonexistent atmosphere with radiation as your only means of shedding heat, heat sinks are not very efficient and TECs get heat saturated.

I definitely recall, but cannot recover, a description of a spacebound generator which used a Stirling engine with high-pressure helium as a working fluid to turn a generator.

Agree with your description of thermocouples to generate power ("convert energy"), with respect to cooling.  When I played with then, I had the same heat dissipation problem.  But, as I mentioned, "efficiency" is not the name of the game in this application.  Overall weight, longevity, and size, is.

I played with a small office Peltier Effect refrigerator (your TECs, I believe) and found the efficiencey was terrible.  Forgot the numbers but many Watts were involved to get the temp down to only about 45°F.  The basic problem was heat from the hot side would get back to the cold side, which was trying to extract heat from the inside of the refrigerator. 

[AZRedhawk44]

I played with a small office Peltier Effect refrigerator (your TECs, I believe) and found the efficiencey was terrible.  Forgot the numbers but many Watts were involved to get the temp down to only about 45°F.  The basic problem was heat from the hot side would get back to the cold side, which was trying to extract heat from the inside of the refrigerator.

I went quite overkill with my experiment.  I had cooling loops running inside the compartment (I was using isopropyl alcohol as refrigerant fluid).  They would pass through a pump and into a liquid channeled heat sink on the cold side of the TEC.  The hot side of the TEC had a separate pump and piping environment that went to an array of radiators with fans attached to them to speed convection with ambient air and move the heat away from the insulated compartment.  I got about a 10 degree temp increase in my hot loop from ambient, and up to about a 40 degree temp decrease in my cold compartment from ambient.  It worked for fermentation, but wouldn't be sufficient for proper refrigeration.  I could just barely knock at freezing with it running full tilt, pulling about 120 watts.  But it effectively acted like a heater in the corner of the house it was in, and I'd estimate that a good 90+% of the power put into it from electricity turned into waste heat. 

Those rocket scientists really had their hands full, getting a TEC to not destroy itself with only passive radiative cooling to control its temperature.  I went through a stack of TECs (granted they were Chinesium Amazon grade TECs) due to cyclic failure.  I always meant to try PWM to throttle the machine, but never got around to it.  I was just using a thermostat relay to engage/disengage the circuit once I hit a target temp.  TECs don't like being turned on and off a lot.

[zahc]

I have always said that solar works best as distributed energy. As long as it's not through force of government, solar roofs on houses, using an existing footprint, make a ton of sense.

If the usual suspects had been pushing that instead of ginormous desert solar farms, with their energy loss through long distribution networks, you could by now probably do solar roofs for the same cost as good quality comp roofs.

Energy loss in distribution is pretty minimal. If you can locate the panels somewhere even minutely better (better sun, cheaper land that letd you throw up a couple more panels, etc.) it will be worth it.

It depends what you optimize for, like always. But I don't think rooftop solar is actually efficient compared to solar farms. If the goal is to generate a fraction of our energy by solar, rooftop is maybe a small part of that but would be a terrible way to go about it.

The idea of rooftop solar is to use space that's otherwise available (your roof) and therefore cheap. The problem is nobody's roof is optimized for the angle, so there's a guaranteed steep loss there vs. a dedicated solar farm (50% is not out of the question). Then there's the quality of panels themselves, maintenance of them (dirty panels can easily lose another 50%), and the conversion efficiency of the small homeowner (read:cost-optimized) inverters is probably a tick worse than a grid scale inverter. So it really makes sense only if you are out of cheap land. You can stick solar panels a lot of places that are otherwise worthless for farming, building or anything else. So I don't think rooftop solar really competes with solar farms at all. They make sense if the individual with the panels can make some money off of them, but I think most people who think they are making money off their solar panels are bad at math and accounting. So that leaves the independence effect of generating your own power. If the panels give you some limited grid independence then they don't have to make money...grid independence is worth money all by itself (look at what gensets cost). I know people that have solar+battery setups as a replacement for a standby genset and they seem very good alternative if you have the space for the panels. I've never penciled out net-metering that makes any sense to me though.

In the category of putting panels in otherwise unused space, I'd rather see more solar farms over parking lots. Parking lots are a tragic waste of space. If you could park under the solar panels it would give shaded parking and walking plus basically open up whole fields that we could blanket with solar panels nearby. It might still be more efficient to park them in the desert though. The US southwest is one of the very best places on the planet for solar.