Mars Ultra-Direct

Last night, I had a new idea. Not technically my own, I grant you, but I thought of a new way to put the pieces together, and cut the costs of a manned Mars mission considerably – especially if amortized over multiple flights. For the purposes of this analysis, I have retained the ‘fixed base’ concept; it remains a significant cost saver, and while it might limit exploration of the planet, it will actually help in what should be a primary mission goal – to work out just how to live on Mars.

The key here is two techniques, neither of which is entirely new to the concept of Mars missions – in-situ resource utilization and on-orbit refuelling. Both of these are critical path technologies, but frankly, both of them are, in my opinion, critical path technologies to any exploration taking place beyond LEO. We’re going to need them sooner or later, and we’re already well ahead of the game in both techniques. (ISRU is essentially Victorian-era chemical engineering, and we’ve been pumping liquids in orbit since Salyut 6.)

Holding with my assumption for a 2031 mission, the first ships leave Earth in 2027, four robotic precursors. Three of them are destined for Mars, each for a different site established as a potential target for long-range human exploration. Essentially, these consist of a group of short-range rovers, with a mission to begin analysis of the local terrain to gather the information required for a manned landing – both in terms of suitability as a safe landing site, and scientific interest, a compromise that will have to be reached by the mission planners before launch. The fourth goes to Phobos and Deimos, and has one mission. Find ice. It’s been theorized that ice exists on the moons of Mars, and we’re going to need it to make this mission work.

If all goes well, then in 2029, the first expedition elements are launched, and here’s the largest single use of launchers for the entire Mars program – four Falcon Heavy boosters, three of them targetted at the selected exploration site, the fourth going to whichever of the two Martian moons proved most suitable for water extraction. Two of the Martian landers will provide the materials required for the base – inflatable hab modules, solar cells, a buggy, equipment, and a water/oxygen extractor. (Knowing that there is water on Mars makes this a lot easier!) The intention is that the astronauts will set the base up themselves upon landing, their key task of the first few weeks – but they’ll have a lander to live in while they work, to use as a ‘construction shack’.

The remaining two landers, one for Mars, and one for, say, Phobos, are the ISRU units for the entire mission. We’re landing them as permanent structures, rather than for a single use. Each will be powered by an RTG, and the one deposited on Mars can serve also as the power supply for the base, with solar cells to supplement. Both have one job – make rocket fuel. SpaceX reports that they are planning on-orbit refuelling of the upper stages of its rockets – so let’s do that in Mars orbit, using fuel refined on Phobos. The escape velocity required to leave that tiny little moon is insignificant – all that is required is a nice big tank and a series of docking thrusters. On Mars, the key is to provide fuel for the lander – I’m assuming a refuelable Red Dragon as a design baseline for the moment. (There is almost certainly also space in the Phobos lander for a trio of LMO communications and observation satellites – if not, that will require another launch.)

Now we have our bases, and we can do things a little better. We only need two Falcon Heavy launchers per mission, now, one of them carrying a Lander/Hab combination to get the crew to Mars, the other carrying the supplies they’ll need for the trip home and for their stay on the surface. Both are possible within the advertised TMI capacity of the Falcon Heavy – you don’t even need to refuel in LEO, but can go all the way in a single hop. (Though I think it wiser to launch the Mars Transfer Vehicle a month early, to allow the astronauts to check that all systems are functioning before leaving Earth for almost three years.) The logistic support ship has two modules – one intended for Martian orbit, to provide the three tons of consumables required for the flight home, the other, a larger module, for the surface, to provide seven tons of supplies – consumables and scientific equipment, primarily. (I’m assuming that half of the weight of the lander will be needed for the lander itself, even on a one-way descent.)

In February 2031, two ships leave LEO – they arrive on Mars in November 2031. Both ships split in half, the two orbital components beginning aerocapture while the lander and surface logistic module commence their descent to the surface. You’d want the logistic module to land first – precise targetting is not required – anywhere within a few miles of the base will suffice. I should note that some consumables will have already been landed in 2029, to serve as a backup should something go wrong.

Touchdown, then in mid-November, 2031. Then the work really begins. On the surface, the astronauts build their base, connecting the already-functional oxygen and water systems first to their lander, then later to the base itself, a collection of modules clustered together for safety. These are intended almost as a ‘wet workshop’ concept – the astronauts will have to install equipment, get everything working, but they have time to do it, and the truly mission-critical elements – fuel, power, oxygen, water, food – will already be operational on landing. If one of the six hab modules is damaged, it isn’t a disaster. In orbit, the hab/upper stage and the orbital logistics module will use aerocapture to enter Mars orbit, a process that will take some weeks, but can be supervised from the surface. Once a stable orbit is attained, the Phobos refinery can transfer fuel to the upper Falcon Heavy stage, providing all the propellent required to get the crew home in 2033. On the surface, they’ll have to transfer fuel to the lander from the surface refinery – the buggy will doubtless be critical here. With a little luck and a lot of work, by New Years’ Day 2032, everything will essentially be ready for the exploration period to begin.

They remain for fourteen more months, leaving Mars in their lander – the same one they used to arrive, refuelled for the flight to LMO – and docking with the refuelled Hab/Logistic/Upper Stage combination. Naturally, they leave the lander behind – though there will be an option of using it for brief excursions to Phobos and Deimos if time permits. (I’d expect at the very least unmanned samplers to be deployed.) This completed, they light their candle and go home.

The next expedition will leave before they arrive – and when the second crew lands, they’ll find everything ready for them, a completed Martian outpost, refineries working to provide the fuel they’ll need for the trip home. Their logistic module will expand their reach – more supplies, potentially more hab modules. The key here is that as much of the mission architecture is rendered reusable as possible, drastically cutting down the number of flights. (Indeed, it’s well within the realm of possibility that the Hab/Upper Stage combination can be reused, refuelled in orbit with a new lander fitted for the trip out, though my current thinking has it as single-use.) The base and equipment are the bulk of the cost – once they’re established, every visit amortizes that cost, and you have the nucleus of a future Martian settlement. I’ve figured four to six expeditions in this first wave, but with the architecture in place, there’s really nothing stopping a far more extended series of missions. If a new base is wanted, then it can always be established later on, just as the first was, though I would instead recommend smaller ‘motels’, a couple of modules as short-duration outposts, with water/oxygen refineries attached. Potentially, for the cost of an additional Falcon Heavy, you could add one for each later mission for limited expense.

With so many of the critical mission systems used for the entire four-mission program, the cost of each is drastically reduced – down to twelve Falcon Heavy launches over ten years. (Fifteen, if the ‘Martian Motel’ option is undertaken.) If as advertised by SpaceX, this is less than a billion dollars in launch costs. Mars Ultra-Direct could be considerably cheaper than Project Apollo as a percentage of GDP, and could almost certainly be undertaken by NASA today, with no budget increase required.

Now all I need is for someone to tell me why this won’t work. (And as an amusing afterthought, I might have finally found a justification for the Asteroid Redirect Mission. If neither of the two Martian moons are suitable for resource extraction, it ought to be possible to find a chunk of ice on a Near-Martian trajectory that is!)