Extracting Power from Nuclear in Space

[PB666]

The problem. Solar electric system need lots of power, however either one has hugely bulk solar electric systems or highly inefficient and heavy fission systems; fusions based power sources are vapor-ware.

http://www.sciencealert.com/graphene-levy-flights-limitless-power-future-electronic-devices

Fission reactors are relatively efficient on land, but require constant servicing because steam generation is rather repair intensive proposition.
Replacing steam generation with thermocouples results in a virtually care free system that generates many times more heat than power.

The above link has an alternative, single layer graphene sheets that can extract heat from ambient air to create current. Whichever system one chooses, this can be used to cool them down and create power.
Its not a perpetual motion machine as it might seem, the heat still has to dissipate somewhere (such as in an ION drive or transformer). It might also be a way to cool ION drives allowing them to work at higher rated powers.
(And thus less weight and less crosssectional area for thrusters)

(-) --------- Ossilator (~)------- Step Up transformer ---------- 10-100KV ------------------->|------------ ION Drive 

Edited November 25, 2017 by PB666

[DDE]

Themocouples and thermionic converters are probably still the way to go. I’m not sure grapheme naosheets would last near a nuclear reactor for any significant amount of time.

[YNM]

I guess it's thermocouples, which is to a good extent yet another PV for different wavelength. If you can get good thermocouples, you also get good PV.

It's really about power density. If one plans interstellar or oort cloud I guess nuclear is the ultimate way. For somewhere in the inner system's space PV is the way.

[TheSaint]

Actually Stirling engines are a very promising application for nuclear power plants in space. They are more efficient than thermocouples, and extremely reliable. Closed-cycle gas turbines would work too, although I would worry that such a high-RPM application would reduce the reliability by comparison.

[DDE]

Of course, the highest-performning nuclear powerplants would involve magnetohydrodynamic converters (which probably trump any other method by efficiency alone, never mind specific power) and core temperatures in the 50000 K region.

[shynung]

2 minutes ago, DDE said:

Of course, the highest-performning nuclear powerplants would involve magnetohydrodynamic converters (which probably trump any other method by efficiency alone, never mind specific power) and core temperatures in the 50000 K region.

When commercial tokamak fusion reactors are ready, I think it's likely to use MHD generators for power extraction.

[DDE]

44 minutes ago, shynung said:

When commercial tokamak fusion reactors are ready, I think it's likely to use MHD generators for power extraction.

I think the tokomaks would (also) end up trying to recapture those pesky 80% of the emissions from the red-hot neutron radiation shielding.

[wumpus]

Nuclear powered electric certainly makes sense for interstellar probes, but I'm not convinced that nuclear thermal power wouldn't be better (I'm not a fan of nuclear thermal power either, just that this has a lot of issues).  One other thing to remember is that "breakeven" issues aren't as critical with thermal exhaust, fusion thermal reactors may have a place in space.

The graphene idea seems to be designed more as an RTG replacement or any other nuclear power system that provides extremely low temperatures.  I'm curious how pebble bed reactors might work in space, although I have even less understanding of the "shelf life" of the pebbles that need to be added and how long they can stay on board spacecraft before use (I'm imaging various pebbles with various radioactive levels stamped with different "use by dates" telling astronauts when to dump them in the core).

There are a few reasons to use electrical methods of propulsion:
1.  You need a ton of delta-v and are willing to wait a long time for the craft to produce it.  In that case electrical methods are ready and have been used for at least a decade (Dawn carried 10km/s of ion power).
2.  800-low thousands of Isp aren't enough (i.e. for an interstellar voyage).  Ion propulsion is already high (there isn't much reason to go higher in the solar system) and other methods go higher (liquid cesium is apparently doing well in the lab), and the likely case the emdrive fails, you can still produce "infinite" Isp by using photons as sources of momentum (you can produce arbitrarily high Isp by firing ions out of cyclotrons).  Don't expect any efficiency from the last two suggestions, but don't forget that the black body radiation emitted from your radiators has momentum, radiating it behind you should make a big difference.
3.  You find a way to scale up ion engines (I'm guessing something like shrinking them down onto a chip and then slapping down a billion of them.  There are hard limits on the power of each engine, but nothing really limits the number you can use [note the chip example wasn't entirely serious.  I think there are density limits as well].  VASIMR works well in the lab, so it is entirely possible that if you solved the problems in storing hydrogen you will be just as happy to use electricity generated by nuclear power to power VASIMR as you would simply heating the hydrogen and using it that way (nuclear thermal has lots of nasty issues, but I still suspect that using it to generate electricity has more).

One last thing, I'd expect the power available from a nuclear thermal rocket to be so extreme that any power needed can be extracted from the waste heat.  In this case graphene or peltiers (or even thermocouples) might be chosen entirely on reliability (or cost if funded by corporations).  I would assume a situation similar to nuclear submarines with power being limited only by the size of the generators (actually it is even worse since the submarine has a heat sink and the spacecraft is limited to black body radiation and any energy used in thrust).

Edited November 25, 2017 by wumpus
again, ctrl enter posted early.

[PB666]

Fission Power systems on Earth are almost entirely steamed based, its a big problem. They are Nuclear Thermal, heat is used to make steam, steam creates work as the differential of gas and liquid.
The problem is that for steam generation the catalyst are the cooling towers and water retention areas. Without these there is no way to develope a differential and what you create is a steam critical instead of prompt critcal bomb.

ION drives have ISPs that exceed 10,000 if that is what you want.
ION drives have 1-efficiency based waste heat generation that ultimately limits output per unit area. We are talking about 100kW per meter which means there is 20kW per meter of waste heat, is no better for VASIMR, in fact its probably worse. For classical ION Drive the thrusters are spread out over a plane where they can radiate from their back sides.

Having said that no-one has really studied the waste heat disposal process. With such a device you could use liquid sodium to transfer waste heat from the ION drives and to a coil where the heat is transfered to the graphene and then used to make additional power.

 

2 hours ago, DDE said:

I think the tokomaks would (also) end up trying to recapture those pesky 80% of the emissions from the red-hot neutron radiation shielding.

Neutrons are useful in converting the hydrogen into deuterium.

[shynung]

6 hours ago, PB666 said:

Fission Power systems on Earth are almost entirely steamed based, its a big problem. They are Nuclear Thermal, heat is used to make steam, steam creates work as the differential of gas and liquid.
The problem is that for steam generation the catalyst are the cooling towers and water retention areas. Without these there is no way to develope a differential and what you create is a steam critical instead of prompt critcal bomb.

Even in space, you'll need the equivalents of cooling towers, since almost all of the power extraction methods described in this thread works on heat flux (temperature difference) - they need a 'hot end' and a 'cool end'. On earth, the cooling towers provide the 'cool end', but in space, radiators are used instead.

[PB666]

2 minutes ago, shynung said:

Even in space, you'll need the equivalents of cooling towers, since almost all of the power extraction methods described in this thread works on heat flux (temperature difference) - they need a 'hot end' and a 'cool end'. On earth, the cooling towers provide the 'cool end', but in space, radiators are used instead.

I think you fission reactor just melted everywhere.

Convection and evaporation transfer heat much, much more efficiently than simple radiation, that is the problem.

Edited November 26, 2017 by PB666

[shynung]

Just now, PB666 said:

I think you fission reactor just melted everywhere.

Not on this particular example. The radiators glow because they are hot - this is to increase heat emission efficiency, reducing needed radiator mass. The actual reactor is buried deep inside the ship.

[shynung]

1 hour ago, PB666 said:

Convection and evaporation transfer heat much, much more efficiently than simple radiation, that is the problem.

In space, conduction and evaporation are infeasible - there are no outside substances (like ocean or atmosphere) to conduct or convect heat to. Radiation is the only way.

The classic way to increase radiation efficiency is to heat up the radiating component, hence the glowing radiator. And, should this be not enough, additional radiating surfaces can be used, which means bigger radiators or more radiators, often both.

[kerbiloid]

Also we should not forget about the Charged Particles resource widely used in KSPIE.
I.e. when an aneutronic fusion reactor (say, B+H) produces alpha-particles, and they are either used directly as a propellant in magnetic nozzles, or give their energy to the magnetic traps.

Edited November 26, 2017 by kerbiloid

[PB666]

1 hour ago, shynung said:

In space, conduction and evaporation are infeasible - there are no outside substances (like ocean or atmosphere) to conduct or convect heat to. Radiation is the only way.

The classic way to increase radiation efficiency is to heat up the radiating component, hence the glowing radiator. And, should this be not enough, additional radiating surfaces can be used, which means bigger radiators or more radiators, often both.

But it cost energy to move heat up a heat gradient instead of away from it. Heat objects tend to radiate.

This is to say in a typical powerplant design the steam in the reactor has the highest energy density, once it transfer via heat exchanger to the turbine water the steam temperature and pressure fall some, then over the turbine it looses more energy density per mole of water. The water that cools this has a lower heat density.

This is typical so if you have a design that creates a different direction (for example your fusion is being conducted in you radiator) then you need to give details and not wave hands.

 

Edited November 26, 2017 by PB666

[shynung]

1 hour ago, PB666 said:

But it cost energy to move heat up a heat gradient instead of away from it. Heat objects tend to radiate.

This is to say in a typical powerplant design the steam in the reactor has the highest energy density, once it transfer via heat exchanger to the turbine water the steam temperature and pressure fall some, then over the turbine it looses more energy density per mole of water. The water that cools this has a lower heat density.

This is typical so if you have a design that creates a different direction (for example your fusion is being conducted in you radiator) then you need to give details and not wave hands.

The simple solution to this is to keep radiator temperature lower than reactor temperature. This way, heat gradient will naturally move heat to the radiators. Sticking a heat engine (closed-cycle gas turbine, thermoelectric, or anything) in between the reactor and radiator heat loop will make it generate energy.

Here's an example design of a fission thermoelectric reactor, from Children of a Dead Earth:

On this design, reactor temperature is kept no higher than 1688 K, radiator is kept at 1200 K. Heat gradient flows to the radiator, not the other way around. Two coolant loops connect the heat engine to the reactor (left loop in diagram) and the radiators (right loop). The heat engine produces 14.6 MW from the heat flux, reactor coolant pump eats ~500 kW, radiator coolant eats another ~600 kW, net power production is 13.5 MW. Coolant loop does not involve phase change (not boiling/condensing water/steam), to keep design simple and reliable.

Edited November 26, 2017 by shynung

[DDE]

5 hours ago, kerbiloid said:

Also we should not forget about the Charged Particles resource widely used in KSPIE.
I.e. when an aneutronic fusion reactor (say, B+H) produces alpha-particles, and they are either used directly as a propellant in magnetic nozzles, or give their energy to the magnetic traps.

Yes, but note how all of those are really, really not-near-future powerplants; of the fission reactors discussed in this thread, I think only the dusty plasma one produces any charged particles at all.

Which brings us to the bimodal fission fragment rocket I love so much.

[Hypercosmic]

People from Children of a Dead Earth community sometimes mention about some kind of 'nuclear diesel engine'. You have molten enriched radioactive fluid and a piston. The piston compresses the enriched radioactive fluid until it went critical, then the explosion force pushes the piston out, then the cycle repeats.

Totally feasible.

Edited November 26, 2017 by Hypercosmic

[kerbiloid]

16 minutes ago, Hypercosmic said:

You have molten enriched radioactive fluid and a piston. The piston compresses the enriched radioactive fluid until it went critical, then the explosion force pushes the piston out, then the cycle repeats.

Probably they have a lot of cheap single-use screwdrivers.
(Of course, engineers arre expendable.)

[PB666]

6 hours ago, shynung said:

The simple solution to this is to keep radiator temperature lower than reactor temperature. This way, heat gradient will naturally move heat to the radiators. Sticking a heat engine (closed-cycle gas turbine, thermoelectric, or anything) in between the reactor and radiator heat loop will make it generate energy.

Here's an example design of a fission thermoelectric reactor, from Children of a Dead Earth:

On this design, reactor temperature is kept no higher than 1688 K, radiator is kept at 1200 K. Heat gradient flows to the radiator, not the other way around. Two coolant loops connect the heat engine to the reactor (left loop in diagram) and the radiators (right loop). The heat engine produces 14.6 MW from the heat flux, reactor coolant pump eats ~500 kW, radiator coolant eats another ~600 kW, net power production is 13.5 MW. Coolant loop does not involve phase change (not boiling/condensing water/steam), to keep design simple and reliable.

But if the properly compacted graphene generator works your power efficiency would be much higher. Note a power efficiency of 21.6 percent.

[DDE]

2 hours ago, Hypercosmic said:

People from Children of a Dead Earth community sometimes mention about some kind of 'nuclear diesel engine'. You have molten enriched radioactive fluid and a piston. The piston compresses the enriched radioactive fluid until it went critical, then the explosion force pushes the piston out, then the cycle repeats.

Totally feasible.

Is somebody reproducing an April 1st article from the Russian Popular Mechanics? They claimed the Soviets made a Volga run on uranium hexafluoride.

https://www.popmech.ru/vehicles/57056-prokatitsya-s-atomom/

I prefer the wheeled badgers.

https://www.popmech.ru/science/11474-kto-izobrel-koleso-evolyutsiya/

Edited November 26, 2017 by DDE

[PB666]

I should add, 13.5 MW is nothing with regard to ION driven power in space.

4 minutes ago, DDE said:

Is somebody reproducing an April 1st article from the Russian Popular Mechanics? They claimed the Soviets made a Volga run on uranium hexafluoride.

https://www.popmech.ru/vehicles/57056-prokatitsya-s-atomom/

I prefer the wheeled badgers.

Its all greek to me.

[shynung]

12 minutes ago, PB666 said:

But if the properly compacted graphene generator works your power efficiency would be much higher. Note a power efficiency of 21.6 percent.

Sure, but in space, there are multiple kinds of efficiencies, and their requirements often clash.

Power plant efficiency (in terms of how many electric W comes out from a given thermal W) relies on high heat flux. That means a low radiator temperature. Cool-running radiators radiate heat less efficiently than hotter radiators, which means a lower mass efficiency (more radiators for a given thermal W input on the generator).

In space, every gram counts (thank Tsiolkovsky for that), so for a given reactor temperature, allowing less efficiency in power generation by increasing radiator temperature would enable the power generation system to get away with less radiator mass.

[wumpus]

17 hours ago, PB666 said:

Having said that no-one has really studied the waste heat disposal process. With such a device you could use liquid sodium to transfer waste heat from the ION drives and to a coil where the heat is transfered to the graphene and then used to make additional power.

One of the huge advantages of nuclear thermal rockets is that they can do similar cooling tricks as chemical rockets: use the expelled mass for cooling before sending it into the "combustion chamber/reactor" for final heating.  Of course, this leaves nasty cooldown issues, you either have to keep ejecting "fuel" (and killing your effective Isp) or somehow deal with cooling a reactor is still radiating plenty of heat/energy.  I've suggested that a rather likely system involves ejecting all used fuel after each "burn".

Obviously the existence of such a graphene system (or even just a peltier or thermocouple) is likely to be used to extract heat as radiators grow larger and larger.  It isn't just the addition of power, just removing the heat is reason enough to do such a thing.  Just don't expect such cooling to remotely compare to an open cycle system where transferring the heat to ejected mass is the whole idea.

[shynung]

4 minutes ago, wumpus said:

One of the huge advantages of nuclear thermal rockets is that they can do similar cooling tricks as chemical rockets: use the expelled mass for cooling before sending it into the "combustion chamber/reactor" for final heating.  Of course, this leaves nasty cooldown issues, you either have to keep ejecting "fuel" (and killing your effective Isp) or somehow deal with cooling a reactor is still radiating plenty of heat/energy.  I've suggested that a rather likely system involves ejecting all used fuel after each "burn".

Obviously the existence of such a graphene system (or even just a peltier or thermocouple) is likely to be used to extract heat as radiators grow larger and larger.  It isn't just the addition of power, just removing the heat is reason enough to do such a thing.  Just don't expect such cooling to remotely compare to an open cycle system where transferring the heat to ejected mass is the whole idea.

If the reactor is powered down, it doesn't emit as much deadly radiation as it does under full power. If a NTR is powered down post-burn, it can be left still hot - the heat will radiate on its own, given enough time - without the crew needing to worry about deadly radiation, as long as the reactor components can withstand the temperature.

Should a heat engine (thermocouple) be plugged in at this state, the reactor would get cold fast, because while heat flows through the heat engine, the reactor isn't producing additional heat to compensate. Then, once the reactor temperature is the same as the radiator's, the heat engine stops generating power, and the reactor stops cooling.

Also, when you wrote 'fuel', you might meant 'propellant' - that's the stuff spewing out of the rocket nozzle. 'Fuel' in the context of nuclear thermal rocket would be the slug of U235 in the reactor itself.