The Successful Failure

By April 1970, flying to the Moon had begun to look routine — so routine that the television networks declined to broadcast Apollo 13's cheerful evening show from space. Aboard were Jim Lovell, veteran commander; Fred Haise; and Jack Swigert, the backup pilot promoted to the crew three days before launch when his colleague Ken Mattingly was grounded — exposed to German measles, which he never actually caught. Fifty-five hours and fifty-five minutes into the flight, some 330,000 kilometres from Earth, mission control asked Swigert to flip a routine switch stirring the oxygen tanks. Damaged wiring inside tank two — the legacy of a years-old manufacturing and testing error nobody had connected — ignited. The tank exploded, tearing off a panel of the spacecraft's side.

The crew heard a bang and felt the ship shudder; alarms lit the panels. Lovell's report to Houston has entered the language: "Houston, we've had a problem." For minutes the data made no sense — so many readings failing at once that controllers suspected the instruments. Then Lovell looked out of the window and saw the truth venting into space: "We are venting something out into the... into space," he reported. It was their oxygen. The command module Odyssey was dying — losing the gas that fed both the crew's lungs and, through its fuel cells, all their electricity and water. The Moon landing evaporated in those minutes; nobody said so aloud, because a far larger question had replaced it.

The answer to that question was hanging on the ship's nose. The lunar module Aquarius — the spidery lander built to carry two men to the Moon's surface for two days — was intact, with its own oxygen, batteries and engine. Within hours the crew powered it up as a lifeboat, shut the dying Odyssey down to save its last battery power for re-entry, and crowded three men into a cabin designed for two. The mathematics were not encouraging: two days of supplies, four days from home — and no way back but forward, around the Moon.

What followed was not one problem but a queue of them, each fatal, each solved in hours by exhausted people who refused to panic. The trajectory came first: mission control chose the free-return path — burn the lander's engine to bend the ship around the Moon and let lunar gravity sling it home. The crew executed the burn manually, steering by the Earth's position in the window, in a vehicle never designed to push the dead weight of the ship it was rescuing. Rounding the far side of the Moon, they set a record no crew has broken: the farthest human beings have ever been from Earth. Haise and Swigert photographed the surface; Lovell, who had flown to the Moon once before and now lost it a second time forever, reminded them gently that getting home was the photograph that mattered.

Then the air itself turned hostile. Three men breathing in a two-man cabin overwhelmed the lander's round CO2 filters, and the dying command module's square spares did not fit — the literal square peg in a round hole. In Houston, engineers dumped onto a table only the objects known to be aboard and built an adapter out of them: a flight-plan cover, plastic bags, a sock, a hose, and grey tape. The procedure was read up to the crew step by step; Swigert and Haise built their copy; the CO2 levels fell. The "mailbox" is still, half a century on, the first slide in every engineering lecture on improvisation.

Behind the drama ran the quieter epic: power. The re-entry module had to be brought back from the dead with a fraction of its normal energy, and the start-up sequence — normally months in the writing — was invented in three days, largely by the grounded Mattingly and controller-legends in the simulators, tested, rejected, rewritten and tested again before a word of it was read to space. The crew copied it by hand in the freezing cabin — the temperature had fallen near zero, water rationed to a fifth of a litre a day, Haise feverish with an infection — onto whatever paper remained. It ran almost without error.

On 17 April 1970 the crew cast off their faithful lifeboat — "Farewell, Aquarius, and we thank you," said Lovell — strapped into the revived Odyssey, and fell into the atmosphere. Re-entry blackout was expected to last about three minutes; it stretched well past four, the longest silence of the mission, before the parachutes broke out of the clouds on live television and the world, which had spent four days holding its breath from Times Square to the Vatican, exhaled. The crew, frostbitten, dehydrated and lighter by kilos, were safe aboard the recovery ship within the hour. The Moon rocks they never collected were the cheapest price ever paid for a legend.

The investigation that followed traced the explosion to its buried origins — a design change and a pre-flight test procedure that had cooked the tank's wiring insulation years before launch — and its findings rebuilt the spacecraft's innards for every later flight. NASA named the mission a successful failure, and meant it precisely: the failure was in the hardware and the processes behind it; the success was in everything built to survive such a failure — the margins, the simulators, the cross-trained teams, and a culture in which the flight director could tell his controllers, as legend has flight director Gene Kranz saying, that failure was not an option, and be believed because they had rehearsed its avoidance for years.

That is why this story ends a science series. The other stories here are about discovery — seeing what no one had seen. Apollo 13 is about the other half of engineering civilisation: what happens when the discovered, designed, tested machine breaks anyway, off the edge of every checklist. The answer, demonstrated across four April days, is that the real spacecraft was never the metal one. It was the years of drills, the habit of speaking calmly, the teams who trusted each other's arithmetic at 3 a.m., and the institutional decision — made long before launch — that people would be trained deeper than the machines could fail. Every organisation owns a version of that choice, and most discover which version they chose the way Apollo 13 did: fifty-five hours in, with a bang.