He who follows Freedom, let him leave his homeland, and risk his life.
— Adam Mickiewicz, Polish poet, 1832
NRPLUS MEMBER ARTICLE L ast week my wife Hope and I traveled to Boca Chica, Texas, to meet with Elon Musk. While we talked inside the SpaceX onsite headquarters, a mariachi band played outside, providing entertainment for long lines of people queued up to apply for multiple categories of jobs building craft to take humans to Mars. Hundreds were already hired and at work in the complex. Soon there will be thousands.
Musk calls his design the “Starship.” It’s a methane/oxygen-driven, stainless-steel, two-stage-to-orbit rocket with a payload capacity equal to the Saturn V booster that sent Apollo astronauts to the Moon. The Saturn V, however, was expendable, with each unit destroyed in the course of a single use. Starship will be fully reusable, like an airliner, and therefore promises a radical reduction in payload-delivery costs.
Starship has yet to be demonstrated. Yet here was Musk, building not the first experimental ship to prove the concept but, as we witnessed touring the place the next day, a shipyard and a fleet. Is he mad? According to conventional aerospace-industry thinking he certainly is. But there is a method to his madness.
I have known Musk for some two decades now. In 2001, I was among those who helped convince him to make Mars his calling. His plan is based to a significant degree on my own work, which is generally known as the Mars Direct plan. Published in 1990 and elaborated in detail in 1996 in my book The Case for Mars, Mars Direct was a radical break with previous NASA thinking on how human Mars missions might be accomplished. But Musk’s Starship plan is far more radical still.
With the exception of a period in the 1990s when NASA, under the guidance of Mike Griffin, the associate administrator for exploration, did embrace an expanded version of Mars Direct, the space agency has stuck with a paradigm set forth by Wernher von Braun in a number of variations between 1948 and 1969. According to those ideas, orbital stations should first be built, providing platforms for on-orbit construction of giant interplanetary spaceships using advanced propulsion systems, which would travel from Earth orbit (or currently, rather more absurdly, lunar orbit) to Mars orbit. Departing from these orbital motherships, small landing craft could take crews down to the Martian surface to plant the flag, make a few footprints, and then return to orbit after a short stay.
In contrast, both Mars Direct and the Starship plan use direct flights from Earth orbit to the surface of Mars, with direct return from the surface to Earth using methane/oxygen propellant made on the Red Planet from local materials. Both plans shun any need for orbital infrastructure, orbital construction, interplanetary motherships, specialized small landing craft, or advanced propulsion. Both involve long duration stays on Mars from the very first mission. For both, the central purpose of the mission is not to fly to Mars but to accomplish something serious there.
But there is a difference. In Mars Direct, the modest earth-return vehicle and the crew’s habitation module both stage off the booster that delivers them to orbit, landing on the Red Planet with a combined useful-habitation-plus-payload-mass of about 40 tons. In Musk’s plan, a Starship is flown to orbit and then refueled there by six tanker Starships, after which the whole ship is flown to Mars, delivering a useful habitation-plus-payload mass of as much as 200 tons. So, while the Mars Direct plan might send crews of four to six astronauts at a time to the Red Planet, a Starship could accommodate 50 or more.
Musk’s plan offers more mission capability than Mars Direct does, but that capability comes with a price. Specifically, if the crew is to come back, you need to refuel a Starship, which needs about 1,000 tons of propellant. In the Mars Direct plan, the much more modest earth-return vehicle sent to the Red Planet in advance of the crew requires only 100 tons. The Mars surface-power and other base requirements needed to support Starship operations are a factor of ten higher than those needed to implement Mars Direct.
So a large base needs to be built in advance, with several Starships sent one-way to Mars and loaded with lots of base equipment, ten football fields’ worth of solar panels, and robots to set it all up. Not until all that is in place can the first crew carrying Starship arrive. That makes the system suboptimal for exploration. But exploration is not what Musk has in mind.
If Mars Direct may be likened to an evolvable version of the Apollo program, Musk’s plan is like D-Day. He needs a fleet. So he’s creating a shipyard to build a fleet. But why build a fleet before testing even one ship? There are several reasons. The first is that Musk wants to be prepared to take losses. By the time the first Starship is ready for its maiden test flight, he’ll have three or four more already built and on deck, ready to be modified to fix whatever caused the first to fail. Launch, crash, fix, and repeat, until it works, and then keep launching, improving payload and cutting turnaround time, advancing performance, flight by flight, ferociously.
But there is another reason to build a fleet. It’s to make Starships cheap. NASA built five space shuttles over a twelve-year period, each one costing several billion dollars. Musk is creating a shipyard designed to ultimately mass-produce Starships at a rate of 50 or more per year. That may sound crazy, but it is not impossible. In 1944, the United States produced escort aircraft carriers at a rate of one per week. Scores of separate teams worked simultaneously, each on its own part of the ship for a few days before passing the job on to the next team. If Musk set up a similar line with a workforce of 3,000, that would mean labor costs on the order of $6 million per ship, or between $15 to $20 million each, with materials and avionics included.
If he can get costs that low, then once the base on Mars is operational, with a growing industrial and greenhouse agricultural capacity, Starships carrying 100 passengers each could fly to Mars and stay there if necessary to provide housing, at a hardware cost per passenger of less than $200,000. So make the ticket price $300,000 — the net worth of a typical homeowner, or about seven years’ pay for an average American. In colonial times, working stiffs booked passage to America in exchange for seven years’ work. It’s a price many people can pay — and have paid — when they really want to make a move. All that is needed besides is Liberty to welcome the immigrants — if she is there, they will come, and prosper through their creativity.
On this latter point, Musk and I agree. An extraterrestrial settlement is unlikely to be able to produce a profit by export of any material commodity to Earth. The transport costs are simply too great, and so the numbers in business plans based on such concepts just don’t add up. But intellectual property is another matter altogether, as it can be transmitted across interplanetary distances nearly cost-free. Bit for bit, the highest value any data can have is that contained in a patent. A Mars colony will be composed of a very technically adept population in a frontier environment where they will be free to innovate and forced to innovate. It will be like 19th-century America, only much more so, a pressure cooker for invention. As historian Frederick Jackson Turner pointed out in his famous essay “The Significance of the Frontier in American History” (1893), an analogous situation made youthful America the most inventive culture ever, with Yankee Ingenuity bringing the world the blessings of electricity, steamboats, telegraphs, labor-saving machinery, recorded sound, light bulbs, telephones, centrally generated electric power — and shortly after he wrote, airplanes and mass-produced automobiles. So, to meet its needs, hard-driven and bureaucracy-free Martian Ingenuity can be expected to produce revolutionary advances in robotics, artificial intelligence, genetically modified organisms, synthetic biology, and many other fields. These inventions, created to meet the necessities of Mars, could be licensed as patents on Earth, bringing in the income needed to fund those imports of complex systems, which unlike bulk materials like food, fabric, fuel, steel, aluminum, glass, and plastic, may be too difficult to make on Mars for some time to come.
Right now, Musk is laser-focused on creating his shipyard, a task he sees as far more central than that of simply perfecting the Starship. But there are many more problems that Musk will need to solve to make all this work. On-orbit refilling of cryogenic propellant tanks has yet to be demonstrated, and Mars in situ propellant production technology, while well understood, is still not ready for use. Starships returning from Mars will encounter much greater heating loads than will vehicles simply reentering from Earth orbit. The lightweight thermal protection that suffices for one might not work for the other. The rocket-exhaust plumes of the massive Starships could create dangerous craters during landings, forcing Musk to adopt a Mars Direct–type plan, staging smaller vehicles, perhaps mini-Starships, off the Starship in Earth orbit. I believe that this consideration, combined with the very large power requirement for refueling a full-size Starship on the Red Planet, may ultimately force him to develop a miniature version of the Starship. Such a “Mini” could be lifted to Earth orbit by a Starship and then staged off it to complete the mission Mars Direct–style, allowing the Starship to return to Earth to be flown to orbit again within a few days. The Mini could also be launched independently, as a reusable upper stage for SpaceX’s already operational Falcon 9, giving the company a fully reusable medium-lift launch capability as well. Musk prefers a doing everything with a single design. We shall see if he can pull it off.
NASA’s operating budget is more than ten times that of Musk’s SpaceX, which, however, is rapidly passing it by. The space agency’s much delayed heavy-lift launch vehicle, now known as SLS, was a reasonable design for a near-term shuttle-derived booster when it was first proposed in 1988. But it is showing up a generation too late, with less payload-delivery capability than Starship’s, and costing about 50 times as much per flight. NASA says it is engaged in an “all hands on deck” effort to land astronauts on the Moon by 2024, but there is little chance that it will, because it has dreamed up a hyper-complex plan that involves building a lunar orbiting space station first and then employing four launches, five flight elements, and six rendezvous operations per mission. While this approach offers the political benefits of giving as many players as possible a piece of the action, the operability of the plan is very questionable.
NASA’s Mars-mission design is even worse. It involves basing a huge ion-drive Deep Space Transport at the lunar-orbit space station and then flying the DST to another space station, one that, the agency claims, needs to be built in orbit around Mars. Transit times from lunar orbit to Mars orbit for this futuristic system is 300 days each way — nearly twice that the Spirit and Opportunity rovers needed to make the trip to from Earth to the Red Planet beginning in 2003. Moreover, unlike Spirit and Opportunity, the DST would not land.
If you want to either explore or settle Mars, you need to land on Mars. The goal of the DST plan, however, is neither exploration nor settlement. It is expenditure. Rather than offer the simplest and most efficient path to the Red Planet, the DST architecture offers the most complex, in order to provide “rationales” (N.b.: not reasons) for as many new technology-development programs as possible.
Musk’s approach is the opposite. NASA’s program is vendor-driven. His is purpose-driven. He is not concerned with justifying expenditures on a raft of “potentially useful” technologies. He wants to get his program done with the least amount of new development. His attitude is “Show me why I need it.” He may push this too far. As noted, I believe he would be wise to develop a Mini-Starship to reduce the power requirements for making return fuel on Mars. He disagrees. “Show me,” he says. Our conclusions on that point diverge, but I really love the way he thinks.
It’s the kind of thinking that can get us to Mars.