by Tim Fernholz
In 2005, Musk had put an additional $11 million into the company, but the expenses of operating in the Pacific had eaten into that quickly. He had also started funding development of the Falcon 9, asking his team to build a more complex rocket while the simple one was still inoperable. SpaceX had generated some revenue; rocket-buying customers put down a deposit on a launch contract; the balance is paid upon delivery to orbit. So far, SpaceX had collected less than $50 million in revenue and spent almost all of it by 2006.
Thus, NASA’s investment in developing a space taxi program arrived at an extremely opportune time. When SpaceX’s second Falcon 1 conked out mid-flight in 2007, the company had collected $80 million from the US government for completing a number of design reviews, and Musk had put in a further $30 million of his own money, the last of the initial $100 million he had earmarked for his space company. That was critical funding for the company at a time when it still lacked a product. “I think we brought them up from being a little hundred-man company, if that, to what they are today,” Mike Horkachuck, the program manager who was SpaceX’s primary liaison with NASA, said later. “Early on, COTS was what was keeping the lights on in the company.”
SpaceX was dangerously close to following in the footsteps of the predecessors that ran out of money before coming up with an operational space vehicle. Despite the focus on cost, the failures and extra expenses of testing on Omelek had increased the company’s burn rate. The NASA funding helped keep the doors open while they built a new Falcon 1, which would allow them to demonstrate a working product and start selling actual flights.
“Absent that, you could debate whether SpaceX would have survived or not,” says Maser, the company’s president when it won its first major NASA contract. His tenure with Musk’s crew lasted just nine months before corporate headhunters tagged him to become CEO of Rocketdyne, the company that built the engines for the Apollo program and the space shuttle. But in his time at SpaceX, he saw what made it unique. Besides NASA’s funding and support, the highly motivated team and their embrace of risk, there was a final difference, and that was Musk himself. The billionaire’s enormous personal investment in the company was made with total patience. “It all ties back to how much risk you want to take,” Maser told me. Most self-funded rocketeers needed to start earning back their investment within three to five years. Musk was entering year six without a rocket.
This willingness to bear risk in the long term was not normal in the aerospace sector, but it is the idea behind the venture capital approach Musk was taking to SpaceX. By 2008, however, despite the NASA contract, SpaceX still needed another injection of serious capital to get its rocket into the sky. No bucks, no Buck Rogers. Unlike Rocketplane Kistler, which had turned to Wall Street, Musk could look to friendly investors with far bigger appetites for risk. He turned to his former partner at PayPal, Peter Thiel. Thiel had parlayed his own newly minted wealth into new investments, including a bet against the US housing market. He also founded, with other veterans of PayPal’s start-up days, a start-up-backed venture called Founders Fund. It was run by high-level entrepreneurs, for high-level entrepreneurs, inspired by Thiel’s personal investment in a then-nascent social network called Facebook.
Now Musk went to his former cofounders. They said they wanted to invest in revolutionary technology? Well, he’d show them a revolution. If anyone could understand the potential gains from transforming a market—in this case for space access—it would be these men. They had already become wealthy commercializing a government technology. The challenge, of course, was that this next level of investing—the rocket business—was very capital intensive. Thiel’s first investment in Facebook had been $500,000. Musk was now asking Founders Fund to invest $20 million— 10 percent of its current funds—in SpaceX.
Whatever personal enmity had led the team to eject Musk from the CEO suite at PayPal, it had not exhausted their faith in him as an entrepreneur. The Founders Fund manifesto famously complained that “we wanted flying cars, instead we got 140 characters,” an unsubtle dig at Twitter and what they saw as the limited ambitions of other Silicon Valley investors. Given a chance to invest in a rocket ship, they said yes, and the first outside investment in SpaceX was sealed in early 2008. The backing of the US space agency and Silicon Valley allowed Musk to avoid the fate of his competitors.
After their investment, Luke Nosek, a Founders Fund partner who joined the board of SpaceX, told me that SpaceX only had enough money to test the Falcon 1 three more times. If SpaceX could not successfully launch the small rocket, it would exhaust its remaining capital and the confidence of its customers. Musk called the new investment “a precautionary measure to guard against the possibility of flight 3 not reaching orbit.” Further failures would mean the end not only of the rocket company but also of Founders Fund’s then-largest investment. It would also mean financial disaster for Musk, who had virtually exhausted his PayPal earnings by plowing it into new ventures, which by now included electric car company Tesla and electricity provider SolarCity, and a lavish lifestyle. He flirted with bankruptcy, later saying that 2008 was the worst year of his life.
“I knew that would wipe him out,” Nosek said. “I also knew that at the last point, we’d have to ask the hard question: how much would we be willing to spend?”
When the third launch campaign for the Falcon 1 began, in August 2008, Nosek headed out on a camping trip in the Sierra Nevada to avoid the stress of monitoring the launch taking place five thousand miles away in the Pacific. He returned to a mobile phone full of condolence texts. The latest test rocket—which carried a cargo of three satellites and the ashes of James Doohan, the actor who played Scotty on the original Star Trek series—had been destroyed mid-flight. (Another portion of Doohan’s cremains would find rest in space on a later SpaceX flight.)
The failure left just two opportunities for SpaceX to actually fly something into orbit. Otherwise, the company’s dreams—of bigger rockets, of flying astronauts to the space station for NASA, or of settling on Mars—would be just another rich man’s flight of fancy.
“There should be absolutely zero question that SpaceX will prevail in reaching orbit and demonstrating reliable space transport,” Musk wrote on the company blog after the third rocket crashed. “For my part, I will never give up, and I mean never.”
SpaceX’s third attempt to fly the Falcon 1 had been sabotaged by its ideology.
The company had used the mission to debut a new variant of the Merlin engine of which it was so proud. Flying the Merlin in the Falcon 1 would help establish its credentials in front of NASA and everyone else ahead of the Falcon 9’s first flight, planned for the next year. Building each of the company’s rockets around the same engine had saved SpaceX billions in development costs. However, to squeeze more power from the engine, it would need to run hotter, and if they wanted to reuse it, it would need to survive that heat largely unscathed. That would not be possible with the flake-off heat shielding the engineers had turned to in order to keep the engine from melting long enough to get the Falcon 1 off the ground.
Their solution to this problem wasn’t novel. Many vehicles use a radiator to cool a hot engine, but everything is more extreme in space. Their plan was to build tiny channels into the walls of the thrust chamber and rocket nozzle, then run the chilled kerosene that fuels the rocket through them. If it sounds a little crazy to use a flammable liquid to cool metal heated to six thousand degrees Fahrenheit, you’re appreciating the confounding power of physics. The fuel, refined to be ultra-stable and cooled to well below freezing, is able to absorb sufficient fury from the running engine before itself becoming the source of heat as it’s pumped into the combustion chamber.
The new iteration of the engine had performed well on the test stand, and, in its first flight, the engine launched the Falcon 1 without a hitch. The eventual failure came in one of those weird flukes that often accompany new equipment.
The meticulously planned flight program called, as usual, for t
he main engines to shut off once the rocket was at the edge of space. Then latches linking the two stages would release, pneumatic pushers would separate them, and the second stage would turn on its engine and continue on its merry way. But the engineers had failed to take into account the effects of the new cooling system. A little extra fuel and oxygen remained in the engine’s plumbing, delivering an unexpected burp of thrust even after it shut off. This was enough to send the first stage careering forward into the second, which sent both tumbling helplessly off course.
The propulsion team had missed the problem in a classic space engineering mishap, according to Musk: the burp was so slight that it was below the ambient air pressure on the ground at their Texas test site. In the vacuum of space, however, it was significant. Ever the optimist, Musk’s update to his fans noted one critical benefit of the incremental, test-as-we-fly philosophy: “We discovered this transient problem on Falcon 1 rather than Falcon 9.”
Learning from the previous two failures—and harking back to Musk’s conversation in the desert with John Garvey years before, when he told Tom Mueller to build two test stands in case one blew up—SpaceX had brought two Falcon 1 rockets out to the Kwajalein Atoll before the third test in 2008. This time the company would not have to wait another year to fly the rocket again. And the engineers were confident they knew exactly how to fix the problem: simply insert a longer delay between main-engine cutoff and separation. “Between the third and the fourth flight we changed one number, nothing else,” Koenigsmann said.
In September, the Falcon 1 once again rose on the launcher arm on Omelek Island, this time with a dummy satellite on top. The launch team performed a “static fire”—a standard prelaunch procedure in which the rocket’s engines are fired through a full burn while the rocket is held down on the pad by heavy clamps—and replaced an oxidizer line they worried about. Then they ignited the engines for real. Once again, the rocket climbed into the heavens on a column of smoke, but the SpaceX team had to endure agonizing seconds watching the skies and their computer screen as they waited for the two stages to separate.
Changing one number was enough. The fourth time was the charm. Not only did the rocket reach orbit impeccably—a first for a privately developed space vehicle—but a further test, shutting off the engine on the second stage and restarting it again, went swimmingly. Musk, in one of the great humble brags, wrote that the launch was “a great relief for me, who led the overall design of the rocket (not a role I expected to have when starting the company).” He admitted that he “felt a little sheepish” receiving an award from the leading organization of US aerospace engineers the week before without a successful launch to his credit. But the recognition, called the George M. Low Space Transportation Award, was fitting. Low, who was a NASA official during the Apollo program, famously skipped the step of orbiting earth with the first crewed Saturn V rocket, instead sending astronauts to orbit the moon—a very test-as-we-fly kind of guy.
The Falcon 1 was the world’s first privately developed liquid-fueled rocket to reach orbit. It was five years later than Musk had planned, but that didn’t matter: the team could tell their competitors “I told you so” with a cheeky grin.
There was just one problem with the newly operational Falcon 1: no one wanted to buy it. Despite hopes of winning a bevy of customers with the low-cost small rocket, the market just didn’t materialize. SpaceX projected as many as a dozen Falcon 1 flights each year but had sold only a handful by the time it got the rocket working. In the interim, the price of the rocket had risen from the envisioned $6 million to $8 million. This was on the edge of affordability for the kinds of scrappy companies and research groups that had small-satellite plans—but it was the wrong side of the edge. Military strategists still wanted the rapid-launch ability for small satellites, but the Pentagon’s priorities were on counterterror missions in Iraq and Afghanistan, not space battles with near-peer competitors.
“You can’t have a market if you have the desire and no money,” Shotwell told me later. The Falcon 1 would fly just one more time, the next summer, to launch a Malaysian imaging satellite. SpaceX would quietly retire the vehicle the next year, moving its existing customers to the forthcoming Falcon 9. The decision was a sound business move, but it also represented something of a betrayal to the small programs and start-ups that had hoped to fly on the rocket.
Whatever problems Falcon 1 had finding a sustainable market, it clearly served as a useful test bed for its successor. That November, at the McGregor, Texas, test facility, the nine-engined first stage went through a full-duration test firing, converting 500,000 pounds of liquid oxygen and rocket fuel into fire and smoke in less than three minutes. SpaceX began to ship hardware to Cape Canaveral, where it hoped to fly the rocket for the first time in early 2009. NASA’s space taxi contract was by far the largest and most important source of revenue for SpaceX at this point. And whether it was the flow of new hardware to Kennedy Space Center, the successful launch of the Falcon 1, or the fact that it had been hitting its COTS program milestones with regularity, SpaceX received an enormous Christmas present that year: a $1.6 billion contract from NASA to fly a dozen resupply missions to the space station. “I love you guys!” Musk said excitedly when he was informed over the phone.
Orbital, the other company in the program, received an eight-mission contract for $1.9 billion, with a price disparity of more than $100 million per flight, underscoring SpaceX’s low cost. Regardless, NASA was still reluctant to trust Orbital and SpaceX, which despite everything else had still flown only one rocket, with a contract of that magnitude. But the clock was ticking on the space shuttle’s retirement, expected in 2010. The replacements had just two years until they would need to enter service. It was “very awkward” for NASA, Gerstenmaier told me later. “We had no choice,” he said. “If we were going to deliver, we needed to go do the services contract, move out, and move forward.” Pressed by necessity, NASA moved, but the decision inspired sour grapes among the veterans of Rocketplane Kistler, who a year before had been begging for that level of commitment just to stay alive. Timing is everything.
January would bring more than just an infusion of government funding for SpaceX. It would also bring a new administration in Washington after the contentious 2008 election, played out against a financial crisis and recession, put Barack Obama in the White House. His campaign had made a theme out of the perceived antiscience stance of the Bush administration, and his transition team had big plans for NASA. But as they arrived at NASA facilities to take stock of existing programs and prepare for change, they encountered an unexpected challenge: Mike Griffin was not done building his rocket yet.
10
Change Versus More of the Same
The truth is, NASA has always relied on private industry to help design and build the vehicles that carry astronauts to space, from the Mercury capsule that carried John Glenn into orbit nearly fifty years ago to the space shuttle Discovery, currently orbiting overhead.
—President Barack Obama, 2010
When the lanky, youthful senator from Illinois arrived in Washington as president-elect, he famously set the scene as “change versus more of the same.” Each new administration receives a list outlining the biggest financial threats to the US Treasury. Obama’s included the financial crisis, the Iraq and Afghanistan wars, the following year’s national census, and the stressed health-care system. But one item focused on NASA’s problem: the space shuttle was still scheduled to retire the next year, with no obvious replacement.
NASA’s bifurcated spaceflight strategy—to build the agency’s own heavy-lift rockets and exploration spacecraft while funding private companies that would service the space station—was under stress. SpaceX and Orbital were building hardware toward their flight dates under the space taxi program. But the Constellation program, the “Apollo on steroids” concept, was already mired in the usual delays of big NASA programs. The previous president had said that Americans would return to the moon by 2015,
but four years later, significant uncertainty remained about whether or not that could be done.
“It’s not just that the rocket takes longer to build; you’re paying for a standing army of people that are on that project for another year,” one NASA executive explained. “So you’re paying for maybe 10,000 people at $200,000 per year. That’s some cost overrun compared to what you originally thought.”
NASA had already contracted out $7 billion to the Constellation project, and the agency anticipated spending some $230 billion more over two decades. Originally, Griffin’s concept was pragmatism distilled: use proven hardware from the space shuttle and the Apollo program to build two modular rockets—one human-rated, for flying a space capsule called Orion with crew, and another, bigger rocket to launch the enormous weight needed to explore the solar system. But the heritage approach turned out to be easier said than done: NASA decided not to use the space shuttle engines as planned, instead hiring Pratt & Whitney Rocketdyne to build a whole new engine. (Rocketdyne is another long-lived specialist contractor, which built engines for the Saturn V, the space shuttle, and the Delta IV; today it is Aerojet Rocketdyne.) Plans to reuse the heat shielding from the Apollo spacecraft on Orion were scotched when engineers couldn’t figure out how to re-create the material. And the teams of engineers who were working to design the new rocket, engines, and spacecraft in tandem kept running into trouble when a change in one system necessitated adaptations in another.
Auditors fretted about unsolved technological problems, schedule slippage, and unrealistic budget forecasts. They were alarmed when Pratt & Whitney, which was on a fixed-cost contract from the government, gave cost-plus contracts to its subcontractors. Griffin wasn’t impressed with their criticism. “We have organizations like the Government Accounting Office investigating our decisions on a launch architecture,” he mused in 2007. “When I was young, NASA’s word on what the launch architecture needed to be was the word.” That year, auditors estimated that NASA had already burned $4.8 billion on failed attempts to replace the space shuttle. Would Ares and Orion just be throwing even more money into the fire?
