I'd like to set a story in our solar system with Earth and the inner planets becoming uninhabitable and a new planet being required. If human beings decided to build a world by making slight deviations to asteroids in the main asteroid belt, trojan asteroids, and Kuiper belt until they start to coalesce, and continue adding asteroids to this until they have an Earth sized planet, and then terraform the resulting world, how would I estimate how long this is likely to take?

I think it's clear it would take too long if the reason was an imminent comet impact with Earth, but what about planning for the expansion of the sun making Earth uninhabitable? Would a few hundred thousand years be long enough for a new planet to be prepared and made ready for life?

I'm interested in what different factors would affect how long this would take:

  • How much can be provided by the three asteroid sources mentioned - are there any elements that would need to be obtained from elsewhere and increase the expected time?
  • In particular would the asteroids release enough gas while coalescing to provide an atmosphere, or would this need to be redirected from elsewhere?
  • Would the planet need to start off molten or could the time required be reduced by choosing glancing approach angles for the incoming asteroids so they spiral in and create less heat on impact?
  • $\begingroup$ @guildsbounty I've posted on meta to see discussion on whether it's useful to leave this question as is, with your answer explaining that there isn't enough mass. Alternative ideas could then be asked as alternative questions. $\endgroup$ Commented Oct 16, 2014 at 14:03
  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$ Commented Oct 24, 2014 at 2:03
  • 3
    $\begingroup$ If you can move an Earth-mass of asteroids inwards from the Oort cloud to build a new planet, you can much more easily move an Earth-mass of Earth (or Venus, if you're still using Earth) outwards from the Sun until it's back in the habitable zone. $\endgroup$
    – Mike Scott
    Commented Aug 23, 2017 at 12:46

6 Answers 6


From asteroids? Can't do it.

To get Earth, you need

$$5.97*10^{24} kg$$

The total mass of the asteroid belt is estimated to have a max mass of $$3.2*10^{21} kg$$ Adding in the Trojan asteroids, you get an additional $$6.4*10^{20}$$

Add in the maximum estimated mass of the Kuiper belt (which is over 30-55 AU away from the sun, and that poses a whole other problem) and you get an extra appx $$5.97*10^{23} kg$$

Add up all your masses, you get


Which is only 10.06% of the mass of Earth.

If you are going to manage this...you're going to need to go all the way out to the Oort Cloud, which we don't actually know anything about, and is situated anywhere from 2,000 AU, to 50,000 AU away from the sun. They guess there is about 5 Earths worth of matter out there. Just for a sense of scale...that puts the upper limit of its distance away at .79 light year, with its theoretical outer edge at 3.16 light years away

Oh, and to further complicate things...asteroids are generally dead, inert rocks. Very little in the way of gasses on board. And there aren't enough comets actively cruising through the solar system any more to deliver that. again, you'd have to go to the Oort Cloud, or mine other planets for material. (Like Mars, Mars seems to have water)

  • $\begingroup$ Interesting answer (+1 for well thought)...where did you get the estimated Kuiper belt mass from? $\endgroup$
    – Twelfth
    Commented Oct 16, 2014 at 22:53
  • 4
    $\begingroup$ Wikipedia and a bit of math. From the article on the Kuiper Belt: The total mass is estimated to range between a 25th and 10th the mass of the Earth with some estimates placing it at one thirtieth of an Earth mass. I went with the most generous estimate (1/10th Earth Mass) and just multiplied the mass of the Earth by .1 $\endgroup$ Commented Oct 17, 2014 at 12:42
  • $\begingroup$ Can we start with Earth? Luna? Venus? Mars? Pluto? These have a bit of heft to them, after all. $\endgroup$
    – The Nate
    Commented Dec 23, 2015 at 22:17
  • 2
    $\begingroup$ @TheNate If we are going to cannibalize other planets, sure...that's definitely possible. But the question as-asked doesn't include that as possibility. $\endgroup$ Commented Dec 29, 2015 at 20:41
  • 1
    $\begingroup$ @TheNate Fair Game? Yeah, sure...but I think it would be better suited as a different, stand-alone question, rather than trying to expand on this answer in the comments. :) $\endgroup$ Commented Jan 5, 2016 at 16:36

I think you are underestimating the vastness of space with this question...but I think you are also thinking a bit 'earth-centric' on posing this question. Ultimately, Earth is a whole lotta notta as far as usability to mass ratio goes and isn't that ideal.

  1. Objects in the Asteroid belt average around 600'000 miles in distance from one another...about 20x the full circumference of earth. That's a lot of effort to even start getting a few of them together.

  2. The kuiper belt is an amazing distance away. If you took the sun and reduced it to the size of a grapefruit, Pluto would be a microscopic piece of dust some 30 yards off in the distance (the image of space and planets taught to us in school books does a poor job...distorts the size a perception if you think pluto and the sun could ever be represented on a sheet of paper). Might get more mass out of the Kuiper belt if you considered Pluto / Charon part of the kuiper belt? The amount of effort in gathering these objects together at this distance here could relocate the entire human race to a new system and back.

  3. This project spans several,. even hundreds, of generations. Do you really think humanity could undertake a project that required several generations of humans to work as a continuous one, with the benefits to be reaped on some abstract my great ^ 100 grandchild could live? We usually revolt in 1-2 generations, let alone hundreds.

  4. A dyson sphere, or variations (dyson ring?) provide a much better ratio of mass to usable space. Rearranging asteroids into a ring around the sun would make far more sense as far as usable mass than attempting to form it into a planet.


Counter Proposal. Why start from scratch? Why not start with one of Jupiter's or Saturn's moons? I am not sure what the new habitable zone would be (someone more knowledgeable about astrophysics can probably tell you that). But that would give you a bigger start. Both Ganymede and Titan are only slightly smaller (on a cosmic scale at least) than Mars and slightly larger than Mercury.

Or, if your story's technology allows it, start with more than one. Ganymede, Callisto, and Io for mass (along with all the asteroids) and Titan for it's Nitrogen atmosphere. Add enough of the other moons (and Pluto), and you should get something bigger than Mars.


Build a "Von Neuman Probe" which basically "social insect like robot" that does two things:

  • Replicates itself using available matter (for generating required energy) and "as-a-self-building-block" than

  • Gather asteroids together or any solar system material available to them according to plan.

Since their numbers increased exponentially like (2-4-8-16-32-64-128-256-512-1024) only available energy and "matter" matters, not the time...

  • 1
    $\begingroup$ In our solar system (which the question implies restricting itself to) there isn't enough material for this to work. Besides, could you really make a planet out of von Neumann probes? :-) $\endgroup$
    – HDE 226868
    Commented Nov 2, 2014 at 0:21
  • $\begingroup$ Ok you have to leave some matter intact to us :). Or you have to consider eat your probes :). Better yet recycle them back to usable matter which adds their dead-weight to desired planetary body. I don't know for sure but if Einstein was right (Energy = mass x square of speed-of-light) there should be remain enough matter to build planetary bodies including wasted-mater-to-probes plus wasted-matter-energy to grow-feed-work them. $\endgroup$
    – underscore
    Commented Nov 2, 2014 at 0:36
  • $\begingroup$ In a smaller scale microbes and other organism (including humans) do it every day. They gather "matter" (from their surroundings) and energy (mostly from the sun) physically change their surroundings (by moving grain of a sand) replicates themselves (almost) than releases their "matter" by dying. But organisms of planet earth lack of genetically and behaviorally imprinted "common goal" like build 1000 miles wide hole in polar caps. If there was one. I am damn sure it's by done billions years ago. I think the idea can be reduced down to "common goal" and "available energy". $\endgroup$
    – underscore
    Commented Nov 2, 2014 at 0:57
  • $\begingroup$ Your desk might be build of the material remainings of particular "Von Neuman Probe" called tree. So you can use your probes as a building block of anything suitable your needs. $\endgroup$
    – underscore
    Commented Nov 2, 2014 at 1:04
  • $\begingroup$ "Coral Reefs" are another some-thing "build-by-von-neuman-probes-with-themselves" that you can build cities on them or can be seen from space. $\endgroup$
    – underscore
    Commented Nov 2, 2014 at 1:48

There would already be a planet there, but the gravity of Jupiter prevents this. It rips the planet apart before it could happen. This would even happen if you brought in extra matter. All you would have is a slightly thicker asteroid belt.

  • 1
    $\begingroup$ I'd be interested to hear any evidence for this assertion. $\endgroup$ Commented Dec 22, 2015 at 12:06

The key considerations are:

  1. Total materials needed, without getting too picky that's about 5.97∗10^24kg of rock, 1.4 × 10^21kg of water, and 5.1480×10^18kg of volatile gases for an atmosphere. That's a big ask, especially the rocky substrate; I'm not sure there's enough building material in the outer system, including rocky moons, to supply this much material.
  2. Bulk material strength, how solid are the objects you're building this world out of? This dictates the thrust you can apply when moving them and thus how fast you can move the material. Not only is this going to be an unknown until you go out and start playing with comets etc... it's also going to be highly variable with otherwise identical looking bodies being orders of magnitude separated in material properties.
  3. How far does the material need to go? Implicit in this is the need to find a stable location to build in, personally I'd look at retrofitting either an existing planet like Mars or a Jovian Moon, maybe Ganymede. Any effort in the asteroid belt is going to be fraught with unwanted collisions and face disruption by Jupiter's gravitational and magnetic fields when it's at closest approach to the work area. Not that it can't be done there, it's just a hassle.
  4. Combining the answers from 2. and 3. gives you average trip time per packet, now you need to work out both how many packets you need, how big are the bodies you're using for the build?
  5. Lastly how many ships are doing the job? This will tell you how many packets are on the move at any one time. This is in many ways dictated by 2. and 4. since building bigger engines than you can use is pointless but ships need to be big enough to handle the largest loads and you can only build so many ships.
  6. Time to completion = total mass / average packet size x average trip time / number of ships making those trips.

I'm going to go with a total of millennia, at a bare minimum, especially since the Oort cloud is the only remaining source for many of the volatile compounds you're going to need and the Oort cloud is the better part of a light year from the sun at closest approach. Also those bodies are likely to be the most fragile and need the longest to accelerate as well as being the most distant. If you want to look at an existing treatment of world building from sub-planetary bodies have a look at Building Harlequin's Moon by Larry Niven.


You must log in to answer this question.

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