In 1952 vonBraun proposed his Ferry Rocket, with three reusable stages.
http://cdn.geekwire.com/wp-content/u...ThirdStage.jpg
Here's SpaceX upper stage recovery
https://www.nasaspaceflight.com/wp-c...012/01/Z54.jpg
In 1933 Sanger proposed his Silverbird Rocket, that used lift to skip off the upper atmosphere to reach any point on Earth.
http://www.luft46.com/misc/sang3.jpg
Its how a second stage might return to the launch center.
The first stage, is capable of flying back without circling the Earth, with some reduction in performance.
https://pbs.twimg.com/media/CW3fDQpUQAAC33M.png
Of course downrange recovery allows larger payloads,
http://i.stack.imgur.com/nrPia.jpg
For larger payloads, the first stage and second stages are recovered downrange on floating platforms underneath the launch trajectory, and then towed back, or later still, refuelled downrange and 'bounced' back to the launch center.
The third stage, lands after completing its mission.
With a LOX/LNG rocket that consists of three common core boosters and an upper stage, that has a 1,750,000 kg take off weight,
Take Off Weight: 1,750,000 kg
Common Core Booster (x3)
Propellant: 441,028.2 kg
Inert: 20,287.3 kg (including landing propellant)
Upper Stage Booster
Propellant: 298,726.0 kg
Inert: 14,304.0 kg (including landing propellant)
Payload: 52,923.3 kg.
The ideal delta vee of the first two boosters is 2.384 km/sec. The ideal delta vee of the central booster is 5.126 km/sec. The upper stage booster attains orbit and flies back to a landing at the launch center after releasing its payload.
https://www.youtube.com/watch?v=sWFFiubtC3c
Now, propellant fraction to attain a 2.384 km/sec delta vee with a 3.4 km/sec exhaust speed is 0.50403 and the inert fraction is 0.49597 so the take off weight is 40,904.5 kg and the propellant is 20,617.2 kg for each of the first two common core boosters that mass 20,287.3 kg.
The propellant fraction to attain 5.126 km/sec delta vee with a 3.4 km/sec exhaust speed is 0.77855 and the inert fraction is 0.22145 which implies a take off weight of the central common core booster of 91,612.4 kg with 71,325.1 kg of propellant and 20287.3 kg inert weight.
Its easier to have a fuel ship haul propellant to the downrange landing platform than it is to have a tug haul the landing platform back to the launch center.
The three common core boosters require a total of 1,323,084.7 kg of propellant. To recover the two outboard boosters downrange after first stage separation 41,234.4 kg of propellant is used. The third stage uses 298,726.1 kg. The central common core booster uses 71,325.1 kg of propellant to return it to the launch center a few minutes after landing far downrange. A total of 1,621,810.7 kg is used on the upside, and 1,734,370.2 kg overall, and at $0.16 per kg (the cost of LOX and LNG) propellant costs are $277,499.23
$5,243.41 per metric ton!
Now, the booster costs $91,000,000 - and if reused 328 times, the capital costs equal the propellant costs, not counting inflation or interest. If reused 3,280 times - the CAPEX is less than the propellant cost.
The entire system is returned to the launch center and relaunched, in less than 8 hours. So, a single ship can be used 3x per day, and would have a 1,100 day service life, before majour rehaul.
Now, if SpaceX puts a power beaming satellite in Geosynchronous orbit, it can beam power to the launch center, and electrolyse sea water. It takes the hydrogen and combines it with CO2 in the atmosphere, to produce methane, and takes the oxygen and uses that for LOX.
With 2.8 kg of LOX for every 1.0 kg of LNG that means we need to create 456,413.2 kg of LNG and 1,277,957.0 kg of LOX.
4 H2O + energy --- 4 H2 + O2 electrolysis
CO2 + 4 H2 --- CH4 + 2 H2O methane
456,413.2 kg of LNG requires the production of 228,206.6 kg of H2 and 1,227,957.0 kg of LOX. This requires reducing 2,054,859.4 litres of water to hydrogen and 1,825,652.8 kg of oxygen by the application of 8,988,804 kWh of energy every 8 hours. Which is 1.1 GW of continuous power.
Now to compete with LOX and LNG at $0.16 per kg, this power from space must be avaialable at $0.03 per kWh. Which produces the propellants needed for $269,665 - per launch (and return).
A power satellite at GEO, that masses 53 tons at LEO, and produces 22 MW per ton in specific power, translates to 1.2 GW! At $1,209 per kg construction cost for the power satellite, its cost is $64 million. This is approximately $0.053 per peak watt. With zero fuel cost and a 20 year life span, with a 'green bond' discount rate of 4.5% that's $832,507 per year. 1.2 million kW sold at $0.03 per kWh earns $315.5 million per year, and has a NPV of $4.1 billion the day it switches on!
SpaceX could sign a long term contract with Solar City to provide sun-fuel to SpaceX from GEO, at $0.03 per kWh and any Tesla would buy any surplus not used by SpaceX beamed to Tesla super charging stations at $0.03 per kWh as well. A deposit, along with contracts from these two companies, are then used to arrange financing to allow Solar City to hire SpaceX to build the power satellite, launch and operate it. Solar City would develop ground stations that receive the energy, and convert it to chemical fuels. Solar City would partner with Tesla to produce Powerwall energy storage units, with Solar City receivers, to sell energy to consumers, and Tesla customers, at $0.03 per kWh. At these prices a fill up would cost $2.70 -of course at the supercharging centers, power is free.
A Tesla motorcar is charged up every 120 hours with 90 kWh of energy. That's 750 Watts for each car on the road. A 1.2 million KW station supports 1..6 million Tesla Motorcars.
A home unit consumes 1.5 kW on average, so each 800,000 homes are supported per station.
A "Tesla family" with two Tesla's and a Powerwall, consume 3.0 kW and pay $65 per month each of these satellites support 400,000 of these families.
At $4.1 billion value for each satellite that's switched on, Tesla is well ahead with each launch! He has more than enough money to develop Mars.
With a 1220.3 meter diameter concentrator covering 1.17 sq km of area, each satellite generates 1.2 GW of continous power at GEO. A production plant that puts up three per day for each launcher that is operating from SpaceX, has the ability to put up more than 1.2 TW of generating capacity and create more than $4.1 trillion per year!
Four reusable launch vehicles launching a rocket once every two hours, and putting up 12 power satellites per day, put up 4.8 TW per year and in five years, transforms the energy picture of Earth!
Tesla can beam power to wherever its needed. Can produce LNG at $0.16 per kg whenever that's needed, from CO2. With MEMS based chemical processing plastics and other fuels can be produced on demand wherever needed, from atmospheric oxygen. All these capabilities, earning a profit, but setting the stage for widespread use of the core technologies on Mars!