Inside the eight desperate weeks that saved SpaceX from ruin

By Eric Berger

Article intro image
Enlarge / The Falcon 1 rocket ascends toward space on its fourth flight.

They bunked in a double-wide trailer, cramming inside on cots and sleeping bags, as many as a dozen at a time. In the mornings, they feasted on steaming plates of scrambled eggs. At night, beneath some of the darkest skies on Earth, they grilled steaks and wondered if the heavens above were beyond their reach. Kids, most of them, existed alone on a tiny speck of an island in the middle of the Pacific Ocean. It was the middle of nowhere, really.

And they worked. They worked desperately—tinkering, testing, and fixing—hoping that nothing would go wrong this time. Already, their small rocket had failed three times. One more launch anomaly likely meant the end of Space Exploration Technologies.

Three times, in 2006, 2007, and 2008, SpaceX tried to launch a Falcon 1 rocket from Omelek Island in the Pacific Ocean, a coral shelf perhaps a meter above sea level and the size of three soccer fields. Less than two months after the last failure, the money was running out. SpaceX had just one final rocket to launch, with only some spare components left over in its California factory.

“We all knew that the stakes were incredibly high,” Zach Dunn recalled of that feverish period in 2008. This time, the Falcon 1 had to work. And the kids knew it. Barely a year out of graduate school and just 26, Dunn nonetheless was a senior engineer over the rocket’s first stage. “It was tense. There was a lot of pressure.”

Today, it is difficult to imagine the world of aerospace without SpaceX. United Launch Alliance, Arianespace, the Russians, and the Chinese likely would still dominate the launch industry, with their prices a closely guarded secret. A decade ago, these industry titans saw in Elon Musk just another gnat to be swatted aside like so many who had come before. The idea of reusing an orbital rocket to lower the cost of access to space? Laughable. Mars?!? This funny sounding guy from South Africa couldn't even put a small, single-engine rocket into orbit.

This, and more, lay on the line September 28, 2008, when SpaceX sought to finally become the first company to privately develop a rocket that successfully reached orbit.

“That's something that only nations had done before, because the barriers to entry were so high,” Chad Anderson, of the Space Angels investment group, said about privately developing an orbital rocket. “Going from zero to one is really, really difficult. And that’s what SpaceX did. That guy and that company have been swimming upstream, and fighting, for so long to get that ball rolling.”

Ten years later, the industry has undergone a radical restructuring. The titans of aerospace have scrambled to remake themselves, or die, in the new world where Musk is not a pretender but the leading protagonist. Following in SpaceX’s wake, more than 100 private companies around the world are trying to accomplish the same feat with rockets of various sizes.

There is no question who blazed the trail or that it began deep in the Pacific tropics. A few days before the fourth and possibly final attempt to launch the Falcon 1 rocket, Dunn left Omelek by boat, making the trip to “the Kwaj,” or Kwajalein Island. This nearby island was tiny, too, measuring about 4km long and several hundred meters across. But it was a continent compared to Omelek and home to SpaceX's launch control center.

Zach Dunn with the Merlin engine that powered the successful fourth flight of the Falcon 1 rocket.
Enlarge / Zach Dunn with the Merlin engine that powered the successful fourth flight of the Falcon 1 rocket.

On the morning of the launch, Dunn sat on console in a bunker on the Kwaj, monitoring the health of the Merlin-1 engine and the fuel tanks in the first stage. As the rocket took off, he watched, hoping it wouldn’t blow up. Cheers erupted inside the bunker about three minutes into the flight, when the second stage separated from the rocket. Then came an agonizing six minutes when the Falcon’s upper stage engine, named Kestrel, would have to burn, necessary to show potential customers the rocket could put their satellites into a proper orbit. And burn it did.

“When Kestrel shut down, the place just exploded,” Dunn recalled of the bunker. “We went absolutely wild. We were all jumping around. Hugging each other. Screaming. It was a righteous celebration.”

The party in the Pacific would last all night. The next day, SpaceX would begin anew in its efforts to conquer the aerospace world. This time, it would not fail.

The beginning

Hans Koenigsmann joined SpaceX as employee number four. He came to SpaceX in 2002 from another aerospace firm in Southern California, Microcosm. (SpaceX’s dynamic president and chief operating officer, Gwynne Shotwell, would follow him from the same company a couple of months later.) Soon, Koenigsmann began to work on an avionics system for a rocket, the brains that controlled it during flight.

With a mandate from Musk to slash costs, Koenigsmann had to decide what to buy and what to build in-house, and he had to hire a team of engineers. In leaving an established company, the German-born Koenigsmann had taken a big risk, leaving a comfortable job for a completely untested venture whose founder was already talking about Mars.

Koenigsmann did not join SpaceX because he believed in that mission. “I just wanted to develop a rocket with a small team, and not 15,000 people like had been done in the past,” he said. “I wanted to show you could do this with 200.”

Initially, SpaceX wanted to launch from Vandenberg Air Force Base, located a couple hours' drive north of its Southern California headquarters. But the big rocket companies, Boeing and Lockheed Martin among them, used that facility for the launch of their rockets, and the military and its contractors were less than welcoming to the upstart.

So eventually, SpaceX looked west. Far to the west, to an Army base 8,000km away surrounded by vast expanses of blue ocean. Here, there was no other traffic on the range. When SpaceX needed priority, it generally got it. “We weren’t in the shadows of other, bigger companies, so that helped us,” Koenigsmann said.

Even today, he can recall the route to Kwaj from Los Angeles. Take an afternoon or evening flight from LAX to Hawaii. Spend the night there and catch the 7am flight to the Marshall Islands from Gate 14. It made several stops along this once-a-day route, and one of them was the Kwaj.

Once there, SpaceX employees mostly rode bicycles around the island, with its sand, surf, and palm trees. When not spending the night on Omelek, they stayed in Kwajalein’s hotel, called Macy’s—which had nothing to do with the department store. “It sounds a little bit like a tropical island, but it’s an Army version of a tropical island,” Koenigsmann said. “So the hotel was the Army version of a hotel. Every piece of furniture had a United States government number.”

Hans Koenigsmann with his family, and the Falcon 1 rocket, at Omelek.
Enlarge / Hans Koenigsmann with his family, and the Falcon 1 rocket, at Omelek.

It took an hour by boat to reach Omelek Island, where SpaceX built its own launch pad and a hangar for its Falcon 1 rocket. Typically, a rocket would arrive three or four months before its flight and require a lengthy series of processing and tests before launch day.

“It’s pretty unique, and people either liked it or hated it; no one was really neutral,” Koenigsmann said of the spare island life in the Pacific. As for the German, truly half a world away from his home, he liked it there beneath the dark Pacific skies. The island reminded him of why he had come there in the first place. “It gave me a real sense of remoteness and how big this planet really is,” he said.

And it also taught him how damned hard it is to break out of Earth’s gravity well.

Page 2

Founded in 2002, SpaceX reached the launch pad with an orbital rocket just four years later. Mid-morning, on March 24, the Falcon 1 lifted off for its maiden flight from Omelek. Initially, the small rocket rose steadily away from the small island. But after about half a minute, with fuel leaking inside the rocket, the vehicle essentially burnt itself up while climbing. Pieces of the Falcon 1 rocket fell back to Earth.

A few hours later, engineers had boated from the Kwaj over to Omelek to salvage what they could. For the young, successful high achievers who had joined SpaceX with big dreams, this would prove a sobering moment.

“Coming down and picking up the pieces was, personally, something that I struggled with for a couple of months,” Koenigsmann recalled. “I was not used to being not successful. But that is part of the business, and I had to get used to it. You have to fight it very hard because otherwise it will take over.”

Almost one year later, having overcome the demons of the first flight, Koenigsmann and the other engineers returned to Omelek with the second Falcon 1 rocket. This flight proved more promising. The booster reached an altitude of 289km, flying well into space, and followed its prescribed track. After the first stage was spent, the second stage separated. However, a harmonic oscillation began at 265 seconds into the flight and grew increasingly unstable until 474 seconds, when the Kestrel engine shut down early. The team could take some comfort from the first stage performance.

“This mission represents a large step forward for SpaceX and the Falcon 1 launch vehicle,” the company’s Flight Review stated. “Although short of complete success, a significant majority of mission objectives were met from both a programmatic and technical perspective.”

The promising second flight set the stage for the third mission, and the payloads selected for this flight reflected SpaceX’s confidence and that of its customers. NASA, the Department of Defense, and a private company, Celestis, would all fly this time.

SpaceX spent 17 months preparing a third Falcon 1 rocket, upgrading its main engine and testing to ensure that everything would go smoothly for a launch in August, 2008. SpaceX began to staff up, too, anticipating greater days ahead. Dunn officially hired on in 2007 after an internship the year before. After a senior manager above him retired, Dunn found himself responsible for much of the first stage, with a console position inside the bunker on the Kwaj for launches.

Seated in the control room with Koenigsmann and about 10 other operators for the third flight, Dunn tracked the countdown and launch with his head down, looking at a computer screen. Data rolled by, displaying the health of the engines and the pressure in the combustion chamber. Dunn felt both excitement and tension.

Truthfully, it almost overwhelmed him, like an out-of-body experience. Dunn felt as though he lost track of time during the Falcon 1’s ascent, even as his first stage performed what Musk later described as a "picture perfect" flight. However, after the first stage separated from the second stage, its new Merlin engine burned slightly longer than anticipated, a slight residual firing that imparted just enough force to cause the two stages to collide.

“When the anomaly occurred, I had my head down,” Dunn recalled. “I was looking at data. And I could just hear this gasp. I looked up and by that point you could just see that things were not right. It took a little while to internalize. It was incredibly disappointing. The team was devastated around me.” Some of his co-workers, in fact, were crying. (The footage of the launch can be seen here. The moment at which the first stage collides with the second one is breathtaking).

For Koenigsmann, who had never failed much before but was now coming to grips with the reality of spaceflight, the third flight was particularly painful.

“We went home, and Elon basically said, ‘We have one more rocket, let’s launch it as fast as we can,’” Koenigsmann said. “The feeling was, overall, this rocket better work or we are in trouble.”

This week, on the cusp of the 10th anniversary of the fourth flight of the Falcon 1 rocket, Musk acknowledged as much. "If we had not reached orbit on that attempt, SpaceX would not exist," he said. "That was a very tough launch emotionally."

A revolution in waiting

Few realized it at the time, but back on the US mainland, the government had begun to plant seeds for a revolution in spaceflight. But they almost certainly would not have germinated had SpaceX continued to fail.

During George W. Bush’s presidency, NASA had dipped its toes into commercial spaceflight by providing several hundred millions dollars to SpaceX and other private companies to develop rockets and spacecraft to deliver cargo to the International Space Station. But Congress, sensing a threat to existing government programs, had the knives out for budgets that would privatize these kinds of services. The repeated SpaceX failures buttressed their claims that only government vehicles like the space shuttle could do serious spaceflight.

Space policy remained far from a significant issue during the presidential election, but by September, 2008, Lori Garver was already working on pre-transition issues for Barack Obama’s presidential campaign. Later, she would lead his transition team on space policy and become NASA’s deputy administrator. She supported the commercial efforts started by Bush and wanted her boss to accelerate them. But it was hard to champion a rocket, the Falcon 1, that kept failing.

“I'd already explained the importance of commercial space launch development to key Obama science and technology advisors, and we were on the path,” she said. “But had it not been a success, it would have been much more difficult to prioritize commercial cargo and crew.”

The early years of the Obama administration saw a pitched battle fought between the White House, which wanted to privatize key parts of NASA to lower costs, and a Congress reluctant to take steps that would cost traditional space jobs near NASA field centers.

A focused campaign from the legacy aerospace companies fueled the battles over funding in Congress. For several years, those attacks reduced funding for commercial cargo and then crew efforts by 50 percent. But the attacks didn’t come from just outside the space agency, Garver said.

“The internal NASA attempts to derail the program were the most difficult to overcome,” she said. “It was like whack-a-mole in the beginning, with only a handful of us with mallets and hundreds of people popping up with tactics to undermine progress.”

Over time, those tactics varied, but they included efforts to make both crew and cargo programs served by a single provider, under a cost-plus program that would have cost NASA much more. Sometimes, positive data about commercial companies was buried or the potential of the agency’s own Space Launch System rocket exaggerated. Requirements were added to the commercial contracts. The sabotage came from everywhere, Garver said, including parts of NASA’s own administration, center directors, legislative affairs, and even some outside advisory committees.

“If Falcon 1 hadn't made it to orbit on that mission, the forces that tried to push us off that course would likely have been more successful in doing so,” she said.

Flight 4

On launch days in the Pacific, the engineers on the Kwaj woke up well before sunrise. Not that Dunn could sleep. He compared the restlessness that night to final exams. For the SpaceX team, the fourth flight offered a binary test: fail, and go find a less exhilarating other job; succeed and go on to conquer the world.

After a quick bite to eat at the Army hotel, it was a 45-minute bike ride to the control center, and that meant facing the wind. The trades blow through the Marshall Islands consistently from east to west, and, due to the orientation of the Kwaj, that meant clipping through a 20-knot breeze to reach the control center.

An overnight team had taken initial steps to prepare the rocket for a launch, such as loading helium onto the vehicle. But Dunn, Koenigsmann, and the others were on hand in the control center five or six hours before launch to oversee the final fueling operations over on Omelek.

There were always technical issues that needed babysitting, but the countdown proceeded smoothly. After the Omelek crew left the tiny island, this time they did not have to return. By 11:15am local time, just 15 minutes after the original liftoff time, Koenigsmann, the chief engineer, and Tim Buzza, the launch director, cleared the Falcon 1 for flight. A final go decision came from Musk, back in California with the rest of the launch team.

Musk, seated, watches the launch of the fourth flight of the Falcon 1 rocket in California.
Enlarge / Musk, seated, watches the launch of the fourth flight of the Falcon 1 rocket in California.

Then, it launched. “After that time you can’t do anything,” Koenigsmann said. “You’re just watching it. We’re sitting on the console, but it has no bearing on the outcome.”

What the engineers watched was a rocket perform as they dearly hoped it would. The first stage burned beautifully, and there were no issues with stage separation. After the Kestrel engine lit, it, too, made its planned burn. The fairing separated. “It was just the greatest thing in the world at that moment,” Koenigsmann said.

Then they had to watch and wonder about their bird for half an hour. Another burn was planned for 45 minutes later, commonly needed to position satellites for their final orbital insertion. A tracking station on Ascension Island in the Atlantic picked up the Falcon’s signal, and the second burn went smoothly as well. After this point the second stage battery was due to die, but it held out just long enough for the ground control stage at Kwajalein to pick it up again overhead.

“It was amazing to see something come back that you had just launched an hour and 30 minutes before,” Koenigsmann said. “That was a pretty good illustration of what an orbit around the Earth means.”

Page 3

The first successful flight of the Falcon 1 served as a catalyst for the commercial space industry. The privately financed development of new rockets and spacecraft, as well as commercial businesses in space beyond telecommunications, had seen ups and downs over the years but never sustained success. SpaceX’s second successful launch of the Falcon 1 rocket in 2009, when it delivered a commercial satellite into orbit, sealed the deal.

“The impact that their first successful commercial launch had on the space industry cannot be overstated,” said Anderson, who runs an investment group, Space Angels, that closely tracks public and private investment in spaceflight. (It has invested in SpaceX multiple times over the years).

Those first two launches were so significant because SpaceX launched a small satellite commercial customer at a price point well below what any other company was doing: $7 million. Perhaps more significantly, the company published their capabilities and pricing online, with a transparency never seen before.

“It opened a curtain into a dark little corner,” Anderson said. “Before this there were a handful of companies serving the government and commercial launch needs, and it was more of a cartel situation.”

Essentially, if a company wanted to launch a satellite, a senior team might fly to Paris and meet with Arianespace for a sit-down meeting with executives there. After determining the requirements, the satellite company would fly home and wait for the launch company to put these requirements into a proverbial black box and come up with a price.

“To launch a satellite before SpaceX, money had to be no object,” Anderson said. “It could be $90 million, or $170 million, or whatever they happened to want that day. These are incredibly formidable barriers to entry for a new venture.”

Price transparency, and SpaceX’s introduction of the Falcon 9 rocket in 2010 at a price of $60 million, helped bring down these barriers to entry. Over time, this unleashed a wave of innovation in satellites and brought spaceflight into the realm of wider entrepreneurship.

Anderson said that, before SpaceX began flying the Falcon 1 rocket, there were a few dozen privately funded companies, globally, engaged in spaceflight activities. Today, there are 350, and they have raised $15 billion in private capital.

It also eased the way for Garver and other advocates of private spaceflight to convince the Obama White House to champion funding for commercial cargo to the space station and, later, commercial crew for astronauts. NASA also opened up the use of the station to commercial ventures so companies could test concepts in space to see if they worked and might one day deliver a profit.

One of the companies that took advantage of this was Made in Space, a 3D-printing company. Its chief executive, Andrew Rush, attributes the development of the International Space Station as a national lab to the Falcon 1 launch. Lower cost access to space has helped nurture a fragile economy in low-Earth orbit, he said.

“Having that commercial infrastructure is a really huge difference maker for companies like Made in Space,” Rush said. “We didn’t have to go and invent cheap rockets. We can rely on cheap rockets. I would say in retrospect that was definitely a tipping point.”


After that first launch, Dunn and the other control center engineers rode their bikes madly across the Kwaj to meet the team that had fueled the rocket that morning on Omelek and then returned to the bigger island. Eventually, they met up on a beach near the dock. Everyone, Dunn recalled, just began to chant “Orbit.” Over and over again.

This initial success helped land a multibillion dollar contract from NASA to deliver cargo to the space station and ignited a furious decade in which SpaceX has gone from a single engine rocket, to one with nine engines, to one with 27 engines. The company has also developed two spacecraft, the Dragon 1 and 2, and landed dozens of first stages.

Outside of the USSR, from 1957 through 1967, and NASA from 1961 to 1971, it is difficult to find a country or company that has had a more dynamic decade in space than this.

For early employees like Koenigsmann and some of the other brilliant engineers who joined the company at the time, this long-sought-after success helped breed confidence in Musk’s ambition of Mars. Settlement of humans there remains an audacious goal, but it no longer seems an impossible one.

“I must say my vision was too small,” Koenigsmann said. “I realized I should put my goals up a little bit higher. So, what about Mars? Absolutely. And that is something that Elon does very well. He basically broadens people’s horizons and makes it clear why Mars should be our next goal and why we should work as hard as we can to make it.”

About two-thirds of the 30 or so engineers on the Kwaj and Omelek for that first successful flight of the Falcon 1 remain at SpaceX today. An experience like that, following an arc from the devastating failure of the third flight, through the long days and nights in tropical heat, to that final, vindicating success has indelibly marked many of them. It was, Dunn says, truly an adventure.

After the ebullient teams met on the beach at the Kwaj, they soon found their way to one of the island’s two bars. The whole of Kwajalein knew about SpaceX and what it was trying to do. They knew SpaceX had had a hard time before that, and many of the military people there were pulling for the plucky firm. So a lot of the island joined the roaring celebration.

They drank the bar empty that night.