In the late afternoon of May 5, SpaceX’s Elon Musk tweeted, “Starship landing nominal!” Musk is not known for understatement. But seeing that stainless steel behemoth soar was, for many, something more like phenomenal. Over 5 million people watched the spectacle on YouTube, perhaps many with bated breath, as every prior attempt at landing Starship had gone up in flames. Not SN15. This Starship, after having climbed 12 kilometers and then coasted down in a “belly flop” configuration—using its wide silver body as a brake—descended slowly, the force of its Raptor engines offering a soft, safe landing.
Some folks at NASA probably felt a sense of relief. To the astonishment of the space industry, in April, NASA had awarded SpaceX a $2.9 billion contract to modify Starship to serve as the system that’ll take astronauts to the moon. The favorite to win the job wasn’t SpaceX, but the heavyweight “National Team,” consisting of Jeff Bezos’ Blue Origin, veteran aerospace contractors Lockheed Martin, Northrop Grumman, and Draper Labs. The selection was so unexpected that, when word of it was first leaked by the Washington Post, some well-informed observers refused to believe it. Politics suggested the National Team was the safe and sure bet.
Perhaps unsurprisingly, the losing teams (which also included an alliance of small businesses led by Dynetics) promptly protested NASA’s choice, temporarily freezing the program. But, since SpaceX offers the most capability, at less than half the price bid by the others, it will likely prevail. NASA will no doubt take heavy fire from Congress for not playing ball. Politicians believe that NASA exists to serve their need to provide economic benefits to their constituents.1 But the government agency also carries the banner of America’s pioneer spirit. It is a human organization, subject to all the flaws of the system that supports it. But it has its moments. And boy, was this one of them.
It was a milestone moment for Musk, too, of course, who founded SpaceX in 2002, fresh off the sale of his digital payments company PayPal, for no less of a purpose than getting humans to Mars. I know, I know. Entrepreneurs—they’re usually in it for the money, right? But the cynics are wrong about Musk. I was among those who helped convince him to make Mars his calling. If he wanted more money, he knew plenty of easier ways to get it than to start, of all things, a rocket company, a notoriously difficult venture with little chance of success. He was looking to do things of immortal importance. Colonizing Mars (along with electric cars and solar energy) made the cut.
Let me underscore just how transformative, and how profound, Starship could prove to be to our future in space, and to our understanding of life. I’ve been in this business for a decent chunk of time. In the late 1980s, I was on the team at Martin Marietta, now Lockheed Martin, that did the preliminary design for what is now called the Space Launch System, NASA’s flagship vehicle. It was originally devised as a quick and dirty way to create a heavy-lift booster out of the then-operational Space Shuttle system components. Starship is nothing like the Space Launch System. It’s unlike anything NASA has made before. It represents an entirely new concept of space operations, and the impact it very well may have on science is extraordinary.
NASA is still claiming that its Artemis program will use the Space Launch System and the Orion crew capsule to deliver astronauts to the “Gateway,” a yet-to-be-built space station orbiting the moon. From there, the idea is to rendezvous with a Starship in orbit that will ferry astronauts to the surface. NASA may fly a few missions that way. But frankly, this is to avoid the embarrassment of having spent so much time and cash on systems that were never used. Once Starship is operational, logic will drive matters in an entirely new direction.
Essentially, the Space Launch System is just the Space Shuttle, but with the airplane-like orbiter deleted, allowing its mass to be replaced by an upper stage and/or a greatly increased payload. Such a simple variant on the Shuttle should have been flying by the mid-90s, and had it been, we could have seen it serve as an enabling technology for a much more capable space program for the past quarter century.
Musk was looking to do things of immortal importance. Colonizing Mars made the cut.
Unfortunately, this did not happen. Despite the fact that a blue-ribbon committee, headed by Jack Kerrebrock, an eminent professor at the Massachusetts Institute of Technology, recommended in 1993 that such a Shuttle-derived heavy-lift booster be quickly developed—as a way of cutting the number of launches required to create the International Space Station by an order of magnitude—he was overruled by then-Vice President Al Gore. Gore wanted to extend the space-station construction program by decades, involving dozens of Shuttle and Russian Proton launches to use as a vehicle for encouraging friendly relations with (i.e. transferring funds to) the new rulers of post-Soviet Russia.2 So the Space Launch System was delayed two decades, until it was already obsolete.
This sadly left NASA with a generation gap of lost expertise. Responsibility for development shifted to people who had never done anything like it before, so the program limped along well into a new century, with a performance-degraded final design and not even a test flight to show for some 30 years of effort, and more than $20 billion worth of expenditures. In 2005, NASA began developing the Orion capsule to give Americans a way to reach orbit after the retirement of the Shuttle, which was expected to occur (and did) around 2010. This should have been a walk in the park, but somehow, again, NASA, its contractors, and Congress managed to turn this into a multi-decade, $20 billion-plus effort, with just one unmanned test flight, in 2014, to show for the effort. Furthermore, not only the cost, but the mass of Orion had ballooned. Coming in at 26 tons, triple that of the Apollo capsule, it was too heavy for the Space Launch System to deliver into low lunar orbit with the propellant needed to return.
So, the Obama administration came up with the idea of building a space station in high lunar orbit. Instead of going to the moon, astronauts riding Orion could drop by the Lunar Orbit Platform, and enjoy the view. Or maybe they could investigate asteroids which could someday be moved into high lunar orbit using advanced propulsion. Wouldn’t that be cool?
Now, I was no fan of the Trump administration. But they must be given credit for recognizing that this plan, which they inherited, was ridiculous. They cancelled the infeasible Asteroid Redirect Mission, and declared that the lunar orbit platform needed to be a “Gateway” to actually go somewhere. So was born the Artemis program, promising to take America back to the moon (and an American woman forward to the moon), by 2024 no less. To avoid embarrassment, the folks at NASA needed the Space Launch System, Orion, and the Gateway to be used as part of Artemis.
But that plan wasn’t looking so good. The Space Launch System program was only willing to vouch for a launch rate of one per year. This is despite the fact that, over the course of its 30-year program, the more-complex Shuttle had achieved an average launch rate of four per year (and peak launch rates of eight per year). So, if a lunar mission was to launch within a reasonable time-frame, it would need to consist not only of a Space Launch System to send an Orion into orbit, but also several more medium-lift expendable booster launches to deliver a vehicle to the Gateway, so that a crew could take it from there, down to the lunar surface, and back. NASA scraped together around $1 billion for engineering-design studies, and put out a request for mission-architecture proposals to industry to develop lunar-excursion vehicle concepts to meet the needs of such a plan.
Starship would not be limited to operating as a ferry from lunar orbit to the surface: It can open the way to Mars.
In April 2020, NASA awarded preliminary design contracts to three contenders—the Blue Origin-led National Team, the Dynetics team, and SpaceX. The National Team, taking the $579 million lion’s share, proposed an awkward, three-stage expendable lander. It corresponded precisely to the unworkable concept NASA had in mind for its mission plan. The Dynetics team of 25 small businesses, taking $253 million, offered a small single-stage with a drop-tanks lander that though (wisely) diverging somewhat, generally fit the NASA paradigm.3
SpaceX, taking the last $135 million, put forward a radically different concept—Starship. It would be an entirely reusable, two-stage-to-orbit, heavy-lift launch system powered by methane-oxygen engines with a capacity about midway between the Space Launch System and the more powerful Saturn V Apollo moon rocket. Because of Starship’s reusability, it would incur less than 1 percent the cost of either. Those features, by themselves, would be world-changing, but there is more: Starship would be designed to be refueled in low-Earth orbit by tanker Starships, allowing it to proceed further, for example to Mars, where its propulsion system could be refueled by propellant readily made from the Red Planet’s abundant water ice and carbon-dioxide atmosphere.
For the simplest Artemis mission—flying cargo one-way from low-Earth orbit to the lunar surface—Starship comes off well, delivering itself, with its large habitable space and propellant-storage volumes, plus an unmatched 100 tons of cargo, provided it can be refueled with eight tanker flights. One drawback is that, for Starship to land, it would have to have its landing engines moved up, to the top of the vehicle, so its powerful rocket exhaust avoids cratering the surface. The bigger problem is providing all the propellant needed to support Starship operations beyond low-Earth orbit.
It would take at least 10 tanker flights to refuel a Starship operating as a ferry between low lunar orbit and the lunar surface, or 14 if it is forced to make use of the Gateway. These requirements, however, could be reduced by developing technologies to extract oxygen from lunar regolith. Moon rocks are composed of a variety of metal oxides averaging about 50 percent oxygen by weight, and Starship’s propellant combination is 78 percent oxygen. By extracting lunar oxygen (and producing metal in the process) the number of Starship flights needed per mission can be cut threefold, greatly accelerating lunar development as a result.
Unlike the competing concepts, Starship would not be limited to operating as a ferry from lunar orbit to the surface: It can open the way to Mars.4 It was designed from the start to make the human settlement of Mars affordable, which is why Starship meets a much more demanding cost-target than anything a mere exploration program might need. Even for a hefty price, like $300 million per astronaut, NASA would jump at the chance to send its people to Mars to explore. But that flight price wouldn’t work for anyone who might want to volunteer to be a Mars colonist. For Mars settlement to be feasible, Starship’s ticket-price has to be cheap enough for a middle-class person to afford.
Such a person might be able to raise $300,000 by selling his (or her) house, and a working stiff a similar sum by mortgaging labor (as was done in Colonial America). Enabling such a ticket price would require cutting launch- and space-transportation costs by at least three orders of magnitude compared to those prevailing today, possible only by making space systems reusable: A Boeing 737 costs about $100 million and typically carries about 100 passengers—if it were expended after one flight, tickets would cost over $1 million. Only by making Starship reusable can space travel, like air travel, be made affordable.
In February 2020, I travelled with my wife, Hope, to Boca Chica, a small city in Texas with a lot of low-lying land, near the Mexican border, where SpaceX is developing Starship and rapidly expanding. Musk wants to incorporate a town there and call it “Starbase.” A mariachi band was playing outside, entertaining long lines of people waiting to apply for jobs. Hundreds were already at work in the complex. Soon there would be thousands. It was apparent that Musk was not building a ship, he was building a shipyard. In the course of its 30-year Shuttle program, NASA built five Space Shuttles, one every six years on average. On our visit, Musk was gearing up to build Starship prototypes at a rate of one per month, which he’s actually done.
Rather than opting to analyze everything for years or decades before any flight tests, as NASA has done, Musk’s approach is to build, launch, crash, fix problems, then try again. He’s pushed his way through almost the entire flight envelope of Starship’s upper stage system. With the success of the SN15 flight, he is now in position to fly it again and again. Musk is aiming for higher altitudes and increased operational perfection until his team can do it blindfolded. SN16 and SN17, incorporating yet more advances on SN15, are nearly complete.
For a space program to be supported, not by three or four flight vehicles, but by scores of them—and eventually hundreds—is revolutionary. Starship ascents will be counted in rates of flights per week, or even per day. The Shuttle’s average flight rate of four per year, meant that, with a program annual cost of $4 billion per year, the actual cost of a Shuttle flight was a whopping $1 billion. A Starship transorbital railroad, employing 5,000 people, would cost about that much per year. Musk is aiming to manage 200 flights, which is possible with 20 operational Starships each turned around to fly again every 36 days. That would work out to $5 million per flight, 1/200th the cost of the Shuttle with five times its payload, for a thousandfold improvement overall.
The benefits of Starship for both robotic and human exploration are hard to overstate. Mars’ recent arrival, Perseverance, can deliver one ton to the Red Planet’s surface. Starship, with its 100-ton capacity, can land a battalion of robots. These could include many Perseverance-like explorers, and much bigger versions of the Ingenuity helicopter. Smaller rovers armed with high-resolution cameras could map the area, transmit to Earth, and allow millions of citizen scientists to walk the landscape in virtual reality and point the machines toward anything interesting. Construction robots, too, possibly humanoid in form, could build a Mars base capable of converting Martian carbon dioxide and water ice into methane-and-oxygen rocket propellant to store in tanks. With such a set-up, fully supplied in advance, Starships could start sending humans.
Rovers are wonderful, but they cannot resolve the fundamental scientific questions that Mars—once very much like the early Earth—poses to humanity: Is life unique to Earth, or did it appear on Mars, too? If so, did it use the same DNA-RNA information system, or something else? Is life as we know it on Earth what life is, or is it just a particular example drawn from a vast tapestry of possibilities? Finding evidence of past life requires fossil hunting. Perseverance will make a stab at that, but human rockhounds—capable of traveling far over difficult terrain, climbing, digging, working delicately, and intuitively following up on clues—can do the job much better. Finding extant life to determine its nature will require drilling down hundreds of meters to reach underground water where life might still thrive, bringing up samples, culturing them, and subjecting them to analysis. That is light-years beyond the ability of robotic rovers.
But there is more. Starship won’t just give us the ability to send human explorers to Mars, the moon, and other destinations in the inner solar system, it offers us a two-order-of-magnitude increase in overall operational capability to do pretty much anything we want to do in space. That includes not only supporting a muscular program of probes to the outer solar system, and making all sorts of experimental investigations in Earth orbit economical, but enabling the construction of giant space telescopes. Much of our knowledge of physics has come from astronomy. This is so because the universe is the biggest and best lab there is. There is no better place to do astronomy than space. The 2.4-meter Hubble Space Telescope has made extraordinary discoveries. What might we learn once we are able to build 2.4-kilometer telescopes in deep space? The possibilities are literally inconceivable.
Robert Zubrin, an aerospace engineer, is the founder of the Mars Society and the president of Pioneer Astronautics. The 25th Anniversary edition of his book The Case for Mars: The Plan to Settle the Red Planet and Why We Must was recently published by Simon and Schuster. Follow him on Twitter @robert_zubrin.
1. Machay, M. & Steinberg, A. NASA funding in Congress: Money matters. European Journal of Business Science and Technology 6, 5-20 (2020).
2. Logsdon, J.M. & Miller, J.R. US-Russian cooperation in human space flight: Assessing the impacts. NASA.gov (2001).
3. While clearly better on a conceptual level than the National Team offering, the Dynetics design never really had a chance, because the team behind it was not credible enough to be given responsibility for something of such central importance to the space program. Rather, they were given a nice chunk of change in order to assure a wide base of support for Artemis.
4. It would only take two tanker flights to fly an unloaded Starship one-way to Mars, or five to send it there carrying 100 tons of cargo.
Lead image: AleksandrMorrisovich / Shutterstock