Rapid unplanned disassembly

FEW things are as spectacular as a successful rocket launch, but a failed one comes close. On June 28th SpaceX, an upstart rocketry firm founded by Elon Musk, an adventurous technology billionaire (see article), began what was to be its seventh uncrewed cargo flight to the International Space Station (ISS). The success of the first such flight, in May 2012, was big news. It was a vindication of NASA’s decision to rely on the ingenuity of the private sector and the discipline of fixed-price contracts to provide cheap access to orbit. As subsequent lift-offs passed without a hitch, press interest faded, and what had been extraordinary quickly became routine.

All again seemed well as the Falcon 9 rocket roared away from its launch pad this week. But two minutes and 19 seconds into the flight one of its oxygen tanks sprang a leak. The rocket powered on for a few more seconds before disintegrating into a shower of glowing debris (see picture). It is unclear, yet, exactly what went wrong. Mr Musk would say only, in the technical jargon beloved of rocket scientists everywhere, that there had been an “overpressure event” in the liquid oxygen tank of the rocket’s upper stage, and that “[the] data suggests a counterintuitive cause”.

Rocket failures are not unusual. Coaxing what is, in essence, a flimsily built flying bomb into orbit without mishap is a difficult task, and most types of rocket in use have suffered at least one accident. This, though, was the first to strike the Falcon 9, which has seen a total of 19 launches. And it adds to a string of rotten luck that has befallen the ISS of late. Of the past seven cargo launches to it (all uncrewed), three have failed.

The first failure was on October 28th last year, when a resupply rocket built by Orbital Sciences, another American space firm, blew up on the launch pad. The second was on April 28th, when a Russian Soyuz launch vehicle released a Progress capsule into a wildly spinning orbit. Ground controllers tried to make contact with the capsule, but on May 8th it and its 2.4 tonnes of cargo burned up in the atmosphere somewhere west of Chile. To add to the perception that all is not well in the private space-launch industry (even though the technology used on this occasion was completely different, and not intended to get into orbit), Virgin Galactic, an aspiring space-tourism company, suffered the fatal crash of a test flight last October.

The Dragon capsule that was supposed to be launched on June 28th was carrying food for the astronauts, equipment for the ISS and a number of scientific experiments. At a press conference following the accident, NASA was at pains to point out that the station’s six-man crew is not in any immediate danger. A Russian resupply mission is scheduled to take off on July 3rd and a further one, courtesy of Japan, is planned for August 16th.

The accident is, nevertheless, the first serious setback in a long time for SpaceX, a firm with a string of successful missions under its belt and launch prices far below those of its rivals. Those facts have attracted many customers. The flight that has just failed was part of a $1.6 billion deal with NASA for a dozen such missions. SpaceX also has an order book bulging with commercial contracts, and in May it was certified fit to launch satellites for America’s soldiers and spies.

Crews control

Jeff Foust, who edits the Space Review, an industry newsletter, points out that SpaceX will have to put future launches on hold while it investigates the problem. “There was nothing unusual about this flight,” he says. “So it’s not obvious what the problem was. If it was some kind of quality-control defect, for instance, it could take them a while to track it down.” That, in turn, could make some customers restless. “Low launch prices are great,” he says. “But ultimately, sticking to a schedule may matter more.”

The accident may also make some American legislators nervous. With the retirement of the Space Shuttle in 2011, America has no way of getting its own astronauts into orbit. Instead it must rent berths on Russian craft. SpaceX is one of two firms (the other is Boeing, an aerospace giant) that NASA is relying on to remedy that defect, through what is known as the commercial-crew programme. Though restricted to hauling cargo at the moment, Dragon is designed with the intention that it will carry astronauts.

In theory, the first crewed flights to the space station will happen in 2017. But the idea of putting astronauts into low-cost, privately developed spacecraft (rather than reassuringly expensive ones developed in the cosier, old-fashioned way) is unpopular with some in Congress, who are flirting with cutting the $1.2 billion the likes of SpaceX are receiving to develop it. Gwynne Shotwell, SpaceX’s president, has pointed out that a crewed Dragon capsule could have survived the break-up of the rocket, thanks to emergency rockets on board that are designed to blast it clear. But Sunday’s failure is likely to give ammunition to these congressional critics.

There is, though, no obvious alternative. Abandoning the commercial-crew programme would mean continuing to rely on the Russians to fly American astronauts. That would be awkward, says Mr Foust, and not just for reasons of national pride. Co-operation between America’s and Russia’s space agencies is close. But relations between their parent governments have deteriorated in recent months. And Russia’s reputation for reliability in space has taken a knock, too. A string of failures (including a probe to Mars, three global-positioning satellites and two cargo trips to the ISS) led to a reorganisation of its space industry in 2013. Engineers who joined during Soviet days are retiring without being replaced, leading to a loss of institutional memory. The discovery in May, by auditors, of $1.8 billion of “financial violations” in the Russian agency has done little to boost people’s confidence.

Everyone’s copybook has thus been blotted over the past few years. Given the complexity of rockets, and the potentially catastrophic consequences of even quite small glitches, such blots are to be expected, so the coincidence of these events is presumably just that—coincidence. Even coincidences, though, can be useful ammunition for hostile politicians.