== Energy to Launch Server Sky ==

Assume that 3 million 3.8 gram thinsats are launched with an Ariane 5ECA, and that each thinsat produces an average of 4 watts of usable power.  The total power is 15 MW .

The 5ECA, with comsat apogee insertion motors, are a 3 stage vehicle:

|| Stage ||               || Fuel weight || MJ/kg || Energy J ||
||   0   || Solid (both)  ||  554,000 kg || 6 ?   ||  3.3E12  ||
||   1   || LOX/LH2       ||  148,000 kg || 16    ||  2.4E12  ||
||   2   || LOX/LH2       ||   14,000 kg || 16    ||  0.2E12  ||
||   3   || Hydrazine     ||    4,600 kg || 10 ?  ||  - - - - ||
||       ||               ||             || Total ||  5.9E12J ||
||<-2>Stack Weight at Pad ||  777,000 kg ||
||<-2>GEO Satellite Dry Weight|| 4,030 kg ||

The mass ratio - launch weight over satellite dry weight - is 193 .  This is approximately proportional to the exponential of the delta V divided by the effective exhaust velocity (including tanks and engines and the rest of the expended booster dry weight).   

||'''Total Delta V''', m/s  ||  M288 ||  M360 ||  M480 ||  M720 ||  GEO  ||
|| Kourou to Transfer       ||  8669 ||  8848 ||  9051 ||  9289 ||  9956 ||
|| Transfer to Injection    ||  1102 ||  1196 ||  1295 ||  1393 ||  1505 ||
|| Stationkeeping 15y       ||     0 ||     0 ||     0 ||     0 ||   750 ||
|| '''Total'''              ||  9771 || 10044 || 10346 || 10682 || 12211 ||
|| Assumed exhaust velocity ||<-5:>               2320                   ||
|| Mass ratio               ||    68 ||    76 ||    86 ||   100 ||   193 ||
|| Dry weight (kg)          || 11400 || 10200 ||  9030 ||  7700 ||  4030 ||
( more information at EarthOrbits )

The energies given are the chemical energies of the fuels.  If we assume '''20% fuel manufacturing energy efficiency''', relative to generated electricity, then the total "electrical" energy needed to fuel an Ariane is 30E12 Joules.  Thinsats can pay back that energy in 2 million seconds, or about '''3 weeks'''.  Not bad.

=== Possibilities ===

The theoretical energy cost of moving 3 grams to the 12789km m288 server sky orbit is derived from the 8669m/s launch delta V and the 1102 m/s insertion delta V, or 38MJ/kg .  That works out to 1.14E5 Joules per thinsat, which a thinsat makes in 8 hours.  While no existing technology can get there, systems such as the [[ http://launchloop.com/ | launch loop ]] can get within an order of magnitude of that. 

Most of the mass of a server sky satellite is a 70 micron aluminum substrate.  Later, if much of the aluminum and interconnect metal are made from lunar materials, the complex technical materials manufactured and launched from Earth can be reduced to a small fraction of that.  

So, although Server Sky makes energy sense now, it can make vastly more energy sense in the future as technological improvement is driven by volume manufacturing.  Getting on a Moore's Law development curve, and moving towards electromagnetic and lunar launch powered by space energy, will continue making the system cheaper.  We '''must''' get on such a development curve if we expect to bring the developing world up to first-world energy abundance while eliminating carbon fuels and remediating their effects.