Apollo 13. Monday 13 April 1970, 21:07 in Houston. Air-to-ground, flight director and CSM EECOM loops, transcribed by hand. Times are hours, minutes and seconds since launch.
FLIGHT is Gene Kranz, the flight director. CAPCOM is Jack Lousma, the only ground voice the crew hears. EECOM LOOP (1) and (2) are unidentified voices on the CSM EECOM channel, the loop for the spacecraft’s power and life support.
55:54:53 [Master caution and warning.]
55:55:18 TELMU Flight, TELMU.
55:55:19 SWIGERT Okay, Houston, we've had a problem here.
[A second voice overlaps, indistinct.]
55:55:19 FLIGHT Go, TELMU.
55:55:20 TELMU We show the LM overhead hatch is closed and the heater
current looks normal.
55:55:23 FLIGHT Okay.
55:55:24 EECOM LOOP (1) We've got more than a problem.
55:55:26 EECOM LOOP (2) Okay, listen, you guys, we've lost fuel cell 1 and 2 pressure.
55:55:26 GUIDO Flight, Guidance.
55:55:28 FLIGHT Go, Guidance.
55:55:28 GUIDO We've had a hardware restart, I don't know what it was.
55:55:29 CAPCOM This is Houston. Say again, please.
55:55:30 EECOM LOOP (1) We lost O2 tank 2 pressure [Pause.] and temperature.
55:55:32 FLIGHT Okay, GNC, do you want to take a look at it? See if you see
any problems?
55:55:35 LOVELL Houston, we've had a problem. [Pause.] We've had a Main B
bus undervolt.
55:55:38 FLIGHT Roger, we're copying it, CAPCOM. We see a hardware restart.The line you know is not on this tape. It arrived later, in the present tense, and a 1995 film made it stick.
It wasn’t the only edit.
Gene Kranz never said “failure is not an option.” The line was written for the film, and Kranz later borrowed it for the title of his memoir.
The film also has him tell a room to work the problem. On the tape, sixteen minutes after the bang, he tells his controllers to solve it. “Work the problem” is the version our industry kept.
These weren’t edits made in ignorance. The crew’s words had been in NASA’s typewritten transcript since 1970, and one of the film’s screenwriters later said the past perfect just wasn’t as dramatic. The room was another matter. When PBS went looking for Mission Control’s master tapes of the accident in 1994, the boxes were empty. The tapes didn’t surface until 2019, in the National Archives. The film rebuilt that room while its own recordings were missing.
Drama sticks because drama photographs. A raised voice is an event. A level one is the absence of one, and a camera has nothing to point at. Every retelling selects for the drama, not only Hollywood’s. In 1983, NASA had a radio programme called Houston, We Have a Problem.
We run the same edit. The incident we retell is the 3am save and the commander who walked in and took the room. The all-hands thanks whoever stayed up all night. It rarely thanks the team whose rollback was running before anyone got paged, because nothing in that story happens.
The first voice from the spacecraft belongs to a man who wasn’t meant to be on the flight. Jack Swigert was the backup command module pilot. Two days before launch, NASA swapped him in for Ken Mattingly, who had been exposed to German measles. Swigert had rehearsed someone else’s part.
Theatre has a word for Swigert’s job. An understudy learns a part to performance standard and may never play it. When the call comes, there’s no time to learn anything. There’s only what the hours put in.
Musicians work the same arithmetic. A passage that sounds effortless from the stalls was played slowly, then less slowly, then hundreds of times, until it stopped being interesting to the person playing it. Boredom is the target. A passage you still have to think about is one you haven’t learned.
The thinking has to go for a reason. Sian Beilock and her colleagues asked experienced golfers to putt, and experienced soccer players to dribble, while paying close attention to their technique. They did worse than when a second task pulled their attention away. Novices did better with the same instruction. A beginner needs attention on the steps. An expert who puts it there comes apart.
Attention is a budget, and every step that runs on its own hands some of it back. The pianist who no longer thinks about fingering can listen to the room. That’s what the hours buy. The audience hears none of this. It hears ease, and calls it talent.
On a bridge, the steps are the sequence: what’s happening, how severe it is, who’s talking to customers, when the next update lands. A commander still thinking about those is spending attention the incident needs.
I should declare an interest. For years I trained new incident commanders with a scene from that film. Ed Harris, as Kranz, walks into a room of arguing engineers, pushes aside a projector that won’t work, quiets them with a hand, his voice barely lifting, lays out the situation and turns the room towards a decision. I dissected it with hundreds of commanders, and I asked the same two questions every time. What do we notice about Kranz? What does he do when the room gives him opposing options?
The answers were rote. He sets the scene. He takes the dissent. He decides on what he has, and he points everyone in one direction. Nobody ever mentioned that his voice stays calm. Nobody mentioned the projector. They watched a calm man and listed his actions.
I hadn’t always had it either. Years earlier, at a managed service provider, I ran incidents for a client whose technical teams belonged to another provider. I can’t remember what was down that night, some ordinary sev2. I can remember the passenger seat of our car, because that’s where I ran half of it. The other provider’s service delivery manager joined the bridge alone and told me, flatly, that he wouldn’t engage anyone to investigate. I couldn’t engage his teams myself. I asked him again, and again. Eventually I got the client’s operations manager on the line, and things began, slowly, to move.
Between the asks, I muted. My wife watched me fight a comms tool while the clock ran and the frustration came up and over. She’s the only witness to what the bridge didn’t hear. On the bridge, I held. On mute, I paid. The drama was mine. Today I’d be past him to the service delivery executive in minutes. That night I didn’t have the move ready, so my attention went to the frame and none of it to the fault.
The calm came later. My opening line arrived by accident, an easy way into a bridge in any state that pulled focus where I wanted it. The form came long before the words. Then I started teaching, the teaching became rote, and the rote became the way I commanded. Watching the scene never showed my trainees the calm. Teaching it a few hundred times built mine.
Before the video call had buffered the squares of panicked faces, the voices arrived, overlapping, none of them pleasant. About twelve squares. “Howdy folks,” I said, in full Australian drawl. “Can someone give me a rundown of what’s going on?” It’s a question. So is the first job Kranz hands out on that tape.
Someone answered with a recount: where we’d been, where we were now. Someone added to it, then someone else, and the voices settled in turn. The fear doesn’t go away. It never has. But a running bridge leaves it nowhere to sit. Sitreps. Action callouts. Marking actions for the PIR. Interrupting. Inviting the quiet ones in. Every word into the channel. One goal: mitigation. Nobody raised a voice.
Then there’s the other kind of bridge. A sev1, hard down, from a deploy whose consequences testing hadn’t caught. The rollback was underway before the page hit my phone. I asked the same question. This time the answer arrived finished: an ETA for restoration, and a breakout already working a Plan B. Ten minutes later we were back, on the ETA they’d given. Nobody had debated the rollback. That decision was made before the incident, so nobody had to make it during one.
The difference between those three bridges is when the decisions got made. On the first, I hadn’t made them yet, so I spent the bridge in a deadlock. On the second, they lived in me. On the third, they lived in the team, and the work was moving before I arrived.
I used to call it a special moment, watching a new commander do the Ed Harris: someone taking the authority in the room and everyone moving into place. It is special. It’s special because the room needed rescuing. For years I taught the rescue as if it were the goal.
You can’t rehearse the failure. Nobody in Mission Control had drilled for an oxygen tank exploding on the way to the Moon. They had drilled something close. Before Apollo 10, a simulation failed the fuel cells at almost the same point in the flight, and the controllers tried to use the lunar module as a lifeboat and couldn’t power it up in time. They lost the virtual crew. On Apollo 13, the lifeboat worked.
At 56:10:45, Kranz speaks to the whole room. “Okay now, let’s everybody keep cool.” The lunar module is still attached and still good, so there’s a way home if they need one. Protect the command module’s batteries. Don’t lose fuel cell 2. Keep the oxygen working. “Let’s solve the problem, but let’s not make it any worse by guessing.”
There is no fix in that speech. Nobody had one yet, and for the first few minutes the room’s best guess had been faulty instruments. The speech is a frame: what we have, what we protect, how we behave while we find out. It’s the part that could be rehearsed, said aloud at the moment it was needed.
Incident command is not a creative activity. The response can be. Later in the flight, engineers on the ground worked out how to make the command module’s square carbon dioxide canisters work in a lunar module built for round ones, using bags, cardboard and tape. A frame invented on the night spends the attention that improvising needs. The commander plays the scales so the room can solo.
So drill the frame and never the answer. Roles. Cadence. Handover. The escalation you’ll want ready when someone on the bridge says no. And if you train commanders with that film, play them the scene. Then play them 56:10:45.
The reward for doing this well is that nobody can tell it was hard.
The rollback sev1 reached the news sites and barely registered. Leadership scrunched their noses. It had been fixed too fast to be of any real concern, and nothing in the response gave anyone anything to point at. A scrunched nose is what a textbook sev1 earns.
It’s also why rehearsal starves. At a company I worked for, I prepared at least four war games. None of them ran. Each was pulled weeks before it was due, because it had been long enough since the last big sev1 that the people who sign these things off couldn’t feel the need. Nearly every quarter, another one landed on my desk. Rehearsal gets funded by the last disaster. The quieter things get, the less of it happens, and quiet is what rehearsal produces.
I make sure every word of a bridge goes into the channel. Read the log from the rollback bridge and it sounds like the loop: timestamps, short lines, everything already in the past tense. There’s nothing in it a film would keep. Do this right, and someone will have to rewrite it to make it a story.
Nobody on that tape says “we have a problem.” Both of them say “we’ve had a problem.” Past tense is how a practised person reports a crisis. By the time they say it out loud, the problem is already being worked.
Endnote
The keep-cool speech, flight director loop, 56:10:45, transcribed by the author:
“Okay now, let’s everybody keep cool. We’ve got the LM still attached, the LM spacecraft is good, so if we need to get back home we’ve got a LM to do a good portion of it with. Okay? Let’s make sure we don’t do anything that’s going to blow our CSM electrical power with the batteries, or that will cause us to lose the main... the fuel cell number 2. We want to keep the O2 and that kind of stuff working, we’d like to have RCS. But we’ve got the Command Module system, so we’re in good shape if we need to get home. Let’s solve the problem, but let’s not make it any worse by guessing.”
Audio: NASA, restored by Ben Feist for Apollo 13 in Real Time. Transcribed and timed by the author, and checked against the site’s transcript. Telemetry from NASA’s detailed chronology of the accident.
Sources: IEEE Spectrum, "Apollo 13, We Have a Solution" (interviews with Jim Hannigan and Bob Legler); The Washington Post, "'Houston, we have a problem': The amazing history of the iconic Apollo 13 misquote" (2017).
Lastly, for the space buffs: EECOM LOOP (2)'s fuel cell numbers are as spoken. NASA's telemetry shows oxygen flow failing to fuel cells 1 and 3.
