We start with one small (but perfectly formed) planetoid. Just for the sake of discussion, we'll say it's Mercury. This planet is more or less useless to me, being too small and too close to the sun. Fortunately, I've got this big building project in mind...and I need materials.

Not-Mercury is therefore slated for deconstruction. I have massive mobile excavator/refinery complexes that are eating steadily away at the crust, processing the rock into useful components, and transporting them to orbit for use in my Dyson Swarm.

However, the core of the planet is still extremely hot - not as hot as Earth's core, and not hot enough to drive tectonic activity, but certainly hot enough to cause critical components of my excavators to become inconveniently runny. What I'd like to do, then, is have some process in place that can transport the heat away from the depths of the planet, keeping the surface cool enough for my machines to work efficiently.

The population of the planet live in subterranean, multi-level tunnel-cities (similar to Zion in the Matrix series). Ideally these cities should be a part of whatever system we put in place to cool the core, but that's not strictly required. Tech level is advanced enough to allow routine interplanetary travel, but not interstellar; fusion reactors, mass drivers, arcologies, etc, but no FTL, matter replication, or transporters. The system is almost completely bare of Handwavium, so ideally tech should be conceivable today.

So - How can I cool down a planetary core on a timescale of centuries?

  • $\begingroup$ So you are deconstructing a planet with population on it? $\endgroup$ Oct 23, 2020 at 23:55
  • $\begingroup$ The population are there primarily to support the deconstruction work. They live in subterranean arcologies while the excavators peel away the surface. Each time an excavator removes the top-most tunnels, everyone just moves down one level and carries on. $\endgroup$
    – Werrf
    Oct 24, 2020 at 1:37

5 Answers 5


The direct answer to your question: Same way we currently cool space ships: By using a heat exchange to capture and transport heat, and radiating that heat off as IR radiation. A solar shade to limit incoming heat from the sun is also helpful.

However I don't think we'll need a planetary core cooling system. Dismantling a planet can't really be done depth first (i.e., speed run to the core), the excavation will be gradual and distributed (shaving a dozen meters off equally around the sphere every year sort of thing). As this is done, hotter rocks are allowed to radiate into space over time and cool down.

You dig another 1m down. Process everything. Accelerate it into space, and move on to another region. When you come back to this region, it's radiated a lot of energy into space and is cooler now.

As you're digging, you're constantly building new refining and launch infrastructure at lower elevations and then dismantling the old, higher elevation one.

Another massive practical reason for doing this is access and power efficiency. Hauling rock up slopes takes a tonne of energy. Hauling rock along flatish ground is relatively cheap. Your robots can move rock better with wheels along smooth near-horizontal ramps than up cliffs.

I've always pictured this project done by an entirely robotic workforce with humans in orbit controlling their strategic moves, the robots being autonomous in tactical moves. No need for a habitat if no humans step foot on the ground.


Launch hot material into space

Basically, there are only three methods to transfer heat: conduction, convection, and radiation. In space, only the method #3 works. Hot object mush exhaust itself radiating heat into space. While this works, this is very slow. You will have to wait for thousand years for a planet-sized object to cool a fraction of a degree. Building big heat-emitting radiators might help, but those radiators must be also planet-sized.

Conduction and convention methods won't work, because there is no other body to conduct the heat to, and there is no atmosphere. But what if you create a convection (or, rather, direct transfer of overheated gas) out of the planet? This would effectively transfer away all of the excess heat.

Your operation then would go like this:

  1. Digging machines would excavate planetary material, vaporizing it in the process. The process would actually become simpler when you get the hotter core;
  2. Vapor would get ionized and transferred to an orbital launch ionic "cannon", which would accelerate material to the first cosmic velocity (which, for Mercury-sized planet is only about 3 km/s) and launch this plasma into the orbit;
  3. The plasma would form a disk around the planet, and due to its low density would quickly cool down, forming dust;
  4. Your orbital operations would collect the dust and transfer it to some other place in the system where you need it.
  • $\begingroup$ The idea is sound but you might want to spread out your material further. With a planetary disk a good chunk of the heat gets radiated directly back onto the planet and there is only so much mass you can store up there before it starts reheating itself though collisions. Also you'd need to launch and then circularize its orbit. Why not launch it into an elongated solar orbit? Shoot it one or two AU to cool down and then when it falls back inwards your construction factories can simply catch and process it. $\endgroup$
    – mlk
    Oct 23, 2020 at 18:37
  • $\begingroup$ @mlk All depends on how distant we want our planetary ring to be. At proper distance, it would be diffuse enough to cool down quickly and not cause much collisions between particles. Sending the dust into interplanetary space also possible, if the stream can be precise enough not to dissipate out there. $\endgroup$
    – Alexander
    Oct 23, 2020 at 19:04

If you can build a Dyson Swarm, you can use other planetoids. May I suggest using Not-Europa, Not-Ganymede, Not-Calisto, Not-Enceladus, Not-Triton etc.?

They are literally covered in ice, which is kinda cool[pun intended], on top of their rocky cores. They are also around half the radius of Not-Mercury or smaller, and farther from the sun than Not-Mercury[citation needed], which means they won't be so hot.

  • $\begingroup$ A good idea, but Not-Europa, Not-Ganymede, Non-Callisto, and the other Definitely-Not-Jovian moons are being used for other purposes, at least during the early phases of construction. $\endgroup$
    – Werrf
    Oct 23, 2020 at 14:50

I think you're dealing with a non-problem here. You're giving a time scale of centuries.

You can't simply dig down in one spot, it won't take too much distance before your dig collapses. You have to spread your digging out all over the planet. As you dig away material what's underneath is warmer but not greatly so. By the time you get back to that spot it will have cooled naturally.

Besides, your big problem is the waste energy from lifting all that material to space, not the energy of the planetary core.


Use the Sundiver Approach

As in the David Brin book. Excess heat is converted to high frequency laser light and either beamed to the surface machinery or into space for other uses. The power is generated using the temperature difference between the surface and sub-surface layers.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .