Board Thread:Fiction Universe Discussion/@comment-9141454-20150625191138/@comment-47205-20150626142220

For nanobots, problems include waste heat (one nanobot in the depths of a swarm of billions, all working away to dissassemble rock, is going to melt), how exactly you build it (they can't just stick any old atoms together to build new working parts for their copies), and random failures that inevitably result when you have countless moving parts (those nanobot numbers really add up).

But if there is a situation where they are feasible, the number of nanobots at a given time t would be N = N02rt, where N0 is the initial number of nanobots and r is the rate at which it makes copies. To work out the total processed mass, you need to multiply this by the processing rate p and integrate with respect to time, which in this simple model just results in (p/2r[ln2])N02rt.

With your estimates of r = 1 s-1, p = 1.157x10-8 kg/s (1 g/day), we get 9.63 million tonnes processed after one minute from one starting nanobot (mainly because it rapidly increases to 260 nanobots by the end). The model may need tweaking. :P (If anyone wants to try, I suggest starting with logistic growth and diffusion.)

Also, to comment on your suggestions in the original post for compounds that could be mined: you say tridignon deflects electromagnetic waves and is transparent, but these contradict, since transparency is a result of there being minimal scattering of light.

The idea of storing energy in a crystal is fairly reasonable, however. I think the most efficient way for that to work is if it were a naturally-occuring room-temperature superconductor, in which case energy would be stored in the motion of electric supercurrents; depending on the behaviour of environmental magnetic fields during its formation, it could even come pre-charged, so to speak.