This article is a data sources and details article.
There is only one system publicly revealed right now: the so-called "BFR", commonly referred to as "Big Falcon Rocket". This was first revealed in 2016 at a meeting in Guadalajara. As first revealed, the two-stage system was larger than depicted today. In a 2017 presentation, a smaller version 9 m in diameter was revealed, and characterized in more detail. There are some differences between the 2017 presentation still on Spacex's website, and a talk given at the Mars Society Convention in 2018, but these seem fairly minor. The 2017 version was reverse-engineered for realistic performance capabilities estimates in Reference 1. These largely confirm what Spacex has so far revealed. Some people refer to the first stage as "BFR", and to the second stage as "BFS", for "Big Falcon Spaceship".
The first stage has 31 Raptor liquid methane / liquid oxygen engines in a configuration that looks like a scaled-up Falcon first stage core, complete with grid fins and landing legs. It is very clear that this is intended to fly back to the launch site, and land on the launch pad, almost exactly the same way as Falcon-9 and Falcon-Heavy first stages recover today. It must retain a modest propellant allowance for that return and landing, as documented in ref. 1.
The second stage is a lifting-body with an external heat shield, intended for aerobraking flyback, followed by a tail-first retro-propulsive final touchdown. This recovery of the second stage for re-use is new. The 2017 presentation shows 6 engines: 4 vacuum types, 2 sea-level types, all liquid methane / liquid oxygen Raptor engines. The 2018 Mars Society talk says there are 7 engines, but did not give details. Payload was 150 tons in 2017, is now "over 100 tons".
For comparative purposes here, the reverse-engineered 2017 version documented in ref. 1 will be used. Nominal max payload is 150 metric tons, carried entirely internally, much the same as an airplane or a ship would do. There is no payload adapter or custom delivery spacecraft in this design. The second stage literally is the lander.
Key to understanding this vehicle design concept is the propellant allowance for landing, after returning from low Earth orbit. This estimate was documented in ref.1, and must still be aboard at entry, or else the vehicle will crash. This is not a large allowance, the vehicle burns most of its propellants (some 1100 metric tons in total) accelerating from the stage point to low Earth orbit. It needs no refilling in low Earth orbit if used as a transport to orbit. There is plenty of propellant aboard the two stages to accomplish flying to orbit, and flying home again separately. Thus to orbit, that is one launch, one payload delivery.
Because of the generality of the internal payload carriage, and the great size of the vehicle, the same system may be used for missions outside of Earth orbit, given refilling with propellants on-orbit. Thus, there are probably at 3 versions of the second stage design: a passenger-cargo vehicle, an unmanned all-cargo vehicle, and a tanker for refilling on-orbit. The passenger-cargo version was well-covered by the 2017 presentation, and some external views shown of the unmanned all-cargo version with a giant clamshell door. No details of the tanker have yet been published. According to Spacex, all three share the same engine and propellant tankage sections. Only the forward sections differ. This is a design concept approach that makes good engineering sense.
Spacex claims that this two stage system, with refilling on-orbit, can take 150 ton payloads to the moon and to Mars. Payloads on the return voyages are restricted to about 50 tons, they say. These claims are verified as feasible by the reverse-engineering analyses in ref. 1. Spacex has not defined how many tanker flights are required for a full-capacity refill on-orbit, nor have they said how much refill propellant these tankers can carry. On the assumption that each tanker can carry 150 tons, or perhaps a just little bit more, then 6 tanker flights plus what would have been the landing allowance, should be pretty close to the 1100 tons capacity. So for a moon or Mars flight, a total of 7 launches should put 150 tons delivered useful payload mass on the surface of the moon or Mars.
Missions to the Moon
Spacex claims that, with reduced return payload and no refilling on the moon, the vehicle can escape the moon and return home to an aerobrake/retro-propulsive landing on Earth. The data in ref.1 confirm the feasibility of doing that mission without refilling on the moon. There is insufficient propellant capacity to recover in Earth orbit. The aerobrake entry is the main means of decelerating at Earth, with retro-propulsive touchdown. If realized in practice, this system could bring a permanently-manned base or presence on the surface of the moon within easy reach. This would be 7 launches to bring 150 delivered useful tons per landing on the moon, plus the vehicle is large and spacious enough to serve as habitable space while there.
Missions to Mars
It is fairly common knowledge that missions to Mars are far more demanding than missions to the moon. A refilled-on-orbit second stage passenger-cargo vehicle (with full 150 ton payload) will expend most of its propellant load just getting from Earth orbit onto the min-energy interplanetary trajectory to Mars, called a Hohmann transfer orbit. It will need the remainder to land on Mars after aerobraking entry, direct from the interplanetary trajectory. There is no propellant to stop in Mars orbit, the touchdown allowance (~1 km/s dV) is simply insufficient for that (>1.6 km/s dV).
The calculations of ref. 1 confirm feasibility of these Spacex claims. Without refilling on Mars, the vehicle is stranded there, forever. To fly a faster trajectory than Hohmann min energy requires carrying a smaller payload, though. Ref. 1 found little or no such excess capability for faster trips at full payload.
Given the infrastructure to create propellant-quality liquid methane and liquid oxygen on Mars, in sufficient quantities within reasonable times, the vehicle could then be fully refilled with the full 1100 ton capacity. Reducing payload to 50 tons, the vehicle then has the capacity to escape from Mars directly onto the min-energy interplanetary trajectory home. It cannot stop in Earth orbit, it only has the propellant to make a free-return aerobrake entry, followed by the final retro-propulsive touchdown.
These vehicles are large and spacious enough to serve as habitation spaces while on Mars. As with the moon missions, it is 7 launches total to send one vehicle to Mars, and deliver 150 tons of useful payload to the surface. They are stranded there permanently, unless propellants can be produced on Mars to refill them.
Cost-Effectiveness for Comparison
This depends upon the per-launch price of a BFR launch. Such is not yet available. Costs are nothing but wild guesses at this time. The published price of a Falcon-Heavy is $90 million. BFR is about twice the size of Falcon-Heavy, but will be fully reusable, not just first-stage reusable. As a wild guess, call it $100 - 200 million per launch. That should be somewhere close, if not in, the ballpark.
We do not need to adjust the published payload numbers, because of the internal carriage inside a vehicle that is also the lander. We need no payload adapter deduction, and we need no lander mass fraction effect. That results in the following:
1. Article dated April 17, 2018, titled "Reverse-Engineering the 2017 Version of the Spacex BFR", located on this site http://exrocketman.blogspot.com