Nothing About Apollo 13 Was Improvised

In early 1969 a Mission Control team sat in an exercise for Apollo 9. The instructors took away the lunar module engines, so the rescue of the lunar module crew had to be flown with the command module. The team solved the problem. The exercise was passed, with nothing to fault.

In the debrief the training boss asked a question anyway. Why did you leave the lunar module powered up? Don’t you recognise how important resources are in case you have some trouble?

Gene Kranz, at the time one of the flight directors in Houston, recounted the moment in NASA’s oral history project in 1999, and admitted that the significance escaped them there and then. They found the question fair and started writing procedures: how the command module could be temporarily vacated and the lunar module run as a lifeboat. A little over a year later, on the evening of 13 April 1970, that was the first thing Mission Control reached for.

The Line Nobody Said

The story told instead is the one everyone knows. Three men, a ruptured oxygen tank, a team on the ground pulling off the impossible with duct tape and stubbornness, and over all of it the line “failure is not an option”.

Nobody ever said it. Screenwriter Bill Broyles wrote it for the 1995 film after Flight Dynamics Officer Jerry Bostick, asked in an interview about panic, replied that they had calmly laid out all the options and failure was not one of them. Kranz liked the compression enough to make it the title of his memoir. That completes the journey: a description of how the work was done becomes a leadership motto. A line about attitude moves the explanation of the success from the procedure to the character, and anyone who tells Apollo 13 that way takes away nothing they could apply on Monday.

What Was Already There

In the same oral history Kranz describes how the team handled unusual situations. There was nothing, however obscure, they would not look at, work through and write up in full. Then they assigned responsibility for it and filed it away, in his words rather like a bookshelf in a library. And then he says the sentence this piece is about: in desperation, when time is short, you want to go back to something you have known and maybe tested before, rather than trying to invent on the spot.

That is exactly how it went. When the oxygen readings fell, the lifeboat plan was already there. That the whole stack could be manoeuvred on the lunar module’s engine had been flown on Apollo 9. And the famous adapter of socks, cardboard and duct tape that connected the command module’s filter canisters to the lunar module was no flash of insight aboard. It was built on the ground before the crew built it. The New York Times put it plainly as early as the summer of 1970: every makeshift procedure carried out in space was first tried out on earth, and rejected if the simulators showed it to be dangerous or impractical.

The adapter the crew assembled from cardboard and duct tape had already been built and tested on the ground. What looked like a flash of insight was the retrieval of a prepared solution.

None of this diminishes the achievement. It consisted of reaching correctly for prepared material under exhaustion and time pressure, and adapting it to a situation nobody had expected in that form. The lunar module was designed to support two people for two days and had to carry three for four. Anyone who takes from Apollo 13 that good preparation makes adaptation unnecessary has the story upside down. The shelf supplied the starting point; the rest came from people who knew their systems well enough to depart from the filed procedures.

The Part That Really Was Unprepared

The organisation was caught out in exactly one place, and it lies before the launch.

In 1965 the specification for the oxygen tank heaters was revised: on the ground they were now to run on 65 volts instead of 28. Nobody changed the specification for the thermostatic switches to match. For five years nobody noticed, because neither qualification nor acceptance testing called for the switches to be cycled under load. On 27 March 1970 the liquid oxygen could no longer be drained from the tank. The workaround was to warm it with those same heaters until it evaporated and could escape as gas. The switch contacts fused shut in the process, leaving the heaters running for hours. The heater tubes probably reached around 540 degrees Celsius, and the wiring insulation was damaged. Nobody saw it, because the tank’s temperature gauge did not read that high.

The review board under Edgar Cortright turned this into a sentence worth keeping for incident reviews: the accident was not the result of a chance malfunction in a statistical sense, but rather resulted from an unusual combination of mistakes, coupled with a somewhat deficient and unforgiving design.

Two Kinds of Memory in One Agency

That leaves two stories side by side, and both belong to the same organisation.

One half had learned how to store experience. After the Apollo 1 fire of January 1967, which killed three astronauts, a discipline took hold in Mission Control that turned every exercise into a procedure and filed that procedure away, even when nobody could say what it might one day be good for.

The other half lost a specification change across five years, two suppliers and several stages of testing. Not through carelessness: every individual check did what it was meant to do. None of them was built to assemble the history of a component. That is precisely what the report recommends at the end, namely that anomaly reviews should in future include the complete history of the components involved. Organisational memory, on this reading, is not a property an organisation either has or lacks. It is a procedure, and it can run impeccably in one part of the house while being absent in another.

The Exercise That Was Worth Something

For practice, the most interesting moment in this story is neither the explosion nor the rescue. It is the debrief of early 1969.

That team had passed the exercise. By any ordinary evaluation logic the matter was closed: objective met, tick in the box, next scenario. The value only appeared because someone asked a question that had nothing to do with passing. Not “did you manage it”, but “what did you burn through doing it, and what would you have needed if things had got worse”.

That kind of question is rarely asked in the debriefs of emergency and crisis exercises. More often the script is built so that the exercise succeeds, because it is meant to demonstrate something to the executive board or the regulator. Design an exercise as evidence and you get evidence. Design it as a search for material and you get material for the shelf. NASA in 1970 had simulators and four control teams working shifts, and hardly anyone has that budget. The question in the debrief is free anyway.

The film ends with the parachutes over the Pacific. The story that gives organisations something to work with begins five years earlier, with a changed voltage specification nobody followed through, and a year before launch, with a question about an exercise that had gone well.

Sources

• Report of Apollo 13 Review Board – Findings, Determinations and Recommendations, NASA, 15 June 1970 (main source)
• Eugene F. Kranz – Oral History Transcript, NASA Johnson Space Center Oral History Project, 28 April 1999
• Eugene F. Kranz – Failure Is Not an Option: Mission Control from Mercury to Apollo 13 and Beyond, Simon & Schuster 2000
• Deliverance from Disaster, The Connecting Link Vol. VIII No. 2, Singer-General Precision, Summer 1970