If Liebergot had been able to look inside the two oxygen tanks, he would not have wanted to risk disturbing them. They, together with the rest of the electrical-generating system, were inside Bay 4 of the service module—one of the six compartments that ran the length of the twenty-five-foot module. The interior of Bay 4, a place of silvery insulation and golden wires, was divided into compartments. The three fuel cells were in the forward one; the two hydrogen tanks were in the rear; and in the middle were the oxygen tanks, two silvery spheres, twenty-six inches in diameter and made of a tough nickel-steel alloy called Inconel, which were strong enough to contain the oxygen under nine hundred pounds of pressure per square inch. They had an outer and an inner shell, and the space between was filled with insulation—some of it inflammable. On top of each tank was a capped dome that sealed an opening for pipes and for wires that brought electricity to instruments inside the tank—fans, heaters, and the sensors for the quantity, temperature, and pressure gauges.
What would have alarmed Liebergot if he had been able to look inside the tanks was that the wires in Oxygen Tank No. 2 were largely bare of insulation. This situation, attributable to both imperfect design and human inattention, had existed for more than two weeks—since March, after a ground crew at Cape Kennedy had piped liquid oxygen into the tanks in a countdown demonstration test. The oxygen and the hydrogen were cryogenic, or cooled to a liquid state, in order to keep them at sufficiently low volume so that they could be compactly stored; the temperature of the oxygen at filling time was two hundred and ninety-seven degrees below zero Fahrenheit. When the test was over, the engineers had been unable to get the oxygen out of Tank No. 2. This trouble may have arisen because the cap on top of the tank had been jolted so that pipes and wires inside were loosened—which may also be why Liebergot was now having trouble with his tank readings. In any event, the ground crew had tried to force the oxygen out by turning on the heaters and fans inside the tank; the fans would stir up the oxygen and the heaters would warm it to make it expand, thus expelling it. The heaters were left on for eight hours—a longer period than such heaters had ever been on before—and during that time nobody was aware that the temperature inside the tank was getting higher and higher. The ground crew was not worried, because they knew there was a thermostatic safety switch in the tank’s interior that was supposed to turn the heaters off if the temperature rose above eighty degrees, the safe limit. So confident were the designers of the equipment that although they provided a thermometer to give a temperature reading from inside the tank, this did not register above eighty-five degrees. What the designers had disregarded, though, was that the safety switch they specified was built to operate on the twenty-eight-volt current of the spacecraft’s power supply, and that when the tanks were tested at Cape Kennedy they were powered from a sixty-five-volt supply. The designers had thought the switch would be kept cool during tests by being immersed in the supercooled oxygen. Instead, as Tank No. 2 emptied, the safety switch overheated and failed. As was determined much later by experimentation with similar equipment under similar circumstances, the switch undoubtedly fused shut so that it couldn’t turn off the heaters. The failure could have been discovered had any of the ground crew noticed that the heaters were still drawing current for hours after they should have turned off, and thus were still in operation; apparently, no one looked at the current gauge. The heat might well have gone up to a thousand degrees—enough to burn the insulation off the wires. After that, if electrical equipment inside the tanks was turned on and the wires happened to come close together, a spark could pass between them.
When Liebergot requested the cryogenic stir, Kranz, the Flight Director, said he would like to hold off awhile on relaying the message, because he wanted to give the astronauts time to settle down after their TV show. Kranz, a big-boned, trim man with fair hair cropped so close that from certain angles it was barely visible, often suggested a tough Marine Corps unit commander. He was wearing a flashy, iridescent white vest, in honor of his team of flight controllers, the White Team. In about an hour, Kranz’s team would be handing over control to the Black Team, and already Black Team controllers had begun to arrive and draw up chairs so that they could look over the shoulders of their White counterparts.
Kranz was delaying action on Liebergot’s request until Haise had returned to the command module. To determine whether he had started back yet, Kranz asked the TELMU, who monitored the data on the LM, whether the LM’s hatch was still open. The TELMU thereupon checked the electrical-power readings for the command module, from which the inert LM was drawing the small amount of power it needed. By pressing a combination of buttons in front of him, each of the flight controllers could throw onto a small television screen in front of him any one of two hundred and fifty charts giving data on the spacecraft’s condition; these were prepared by computers on the ground floor of the Operations Wing, where the telemetry was received. There had been a slight drop in the power output, so the TELMU guessed that the LM’s hatch was closed—since the lights that turned on when it opened were now drawing no electricity—and therefore that Haise was on his way back to the command module. Accordingly, Kranz told the CAPCOM to radio to Swigert, the command-module pilot, the message to stir the tanks. Swigert, looking up into the apex of the command module, could see Haise coming back through the short tunnel, and that was about all he could see, for he was hemmed in by over five hundred dials, buttons, knobs, switches, and thumbwheels. Most of them were guarded by little U-shaped wickets, lest an astronaut bump against one inadvertently. Swigert’s movements were gingerly; as the new crew member, he was especially anxious to perform as he should. When he received Liebergot’s message, he pressed four switches to his right. In the Control Room, Liebergot sat forward to get a better look at the screen on his console, which would now show the pressure, quantity, and temperature readings of the tanks.
Nothing much happened for sixteen seconds. Then, inside Oxygen Tank No. 2, an arc of electricity shot between two naked wires. In the next twenty-four seconds, the arc heated the oxygen, and its pressure rose rapidly. Because the hydrogen-tank low-pressure signal had preempted the system, no caution lights flashed, and because Liebergot was concentrating on the readings for the hydrogen tanks, which were on the right side of his television screen, he didn’t notice the rapidly increasing numbers in one of the oxygen-pressure columns, three inches to the left of where he was looking. During the time the pressure in the oxygen tank was increasing, the only person in the Control Room to notice that anything was wrong was William Fenner, the GUIDO, who saw what he called an “event”—an unexpected number—on his console. It signalled what he called a “hardware restart,” which meant that the spacecraft computer had found a problem and was going back over recent events to find out where the trouble lay. It never found out, but the restart provided Kranz with a false trail to follow later.
On the basis of recorded data, of evidence brought back by the astronauts, and of extensive post-mission analyses, it is possible to reconstruct with a fair degree of certainty what happened during this two-minute period. At the end of twenty-four seconds—at eight minutes past nine—the oxygen pressure had blown the dome off the top of the tank. The layer of insulation between the inner and outer shells of the tanks undoubtedly caught fire, with flames, fanned by the rush of escaping oxygen, spewing as from a blowtorch all over the inside of Bay 4 of the service module. The silvery sheets of Mylar insulation—heat-resistant but nevertheless inflammable—lining the inside of the bay probably caught fire, and the resulting gases blew out the bay’s cover, which was one of the six panels making up the service module’s external hull. It was lucky the panel blew out when it did, for if the pressure had been allowed to build up much more, the command module itself, plugging the front end of the service module like a cork, could have blown off instead. Later, in describing what happened, NASA engineers avoided using the word “explosion;” they preferred the more delicate and less dramatic term “tank fai
lure,” and in a sense it was the more accurate expression, inasmuch as the tank did not explode in the way a bomb does but broke open under pressure.
Whether called an explosion or a tank failure, such an event is less noticeable in space than it would be on the ground, where air transmits sound and shock waves. Therefore, none of the astronauts were aware that one of the oxygen tanks had ruptured. Nevertheless, each of them was instantly made aware, in one way or another, that there had been an untoward event. First, Swigert reported over the radio that they seemed to have a problem. His voice was so calm that the CAPCOM, Jack R. Lousma, could not tell which of the astronauts was speaking, and Lousma knew the astronauts well, because he was an astronaut himself. What had disturbed Swigert, as he later recalled it, was not so much the sound of a perceptible bang as the sensation of a sort of shudder that ran through the spacecraft. He could not make a precise distinction, he said, because the borderline between feeling a vibration and hearing it is sometimes imperceptible. What he felt may in fact have been not unlike the disconcerting shudder that first puzzled some of the passengers aboard the Titanic as the ship scraped against an iceberg. Swigert was strapped into his seat, and so was better able to feel the shudder than Lovell, the spacecraft commander. Lovell, who was floating just above his seat, said later he had not felt the shudder but had heard a distinct bang. Lovell’s first thought was that the bang had been made by Haise opening a valve in the lunar module. At thirty-six, Haise still looked like the youthful, irrepressible sort of person who might make a loud noise without warning. However, Haise was at this moment emerging from the tunnel, and Lovell could tell by the look on his face that he, too, had been jolted by something. Far from causing the bang, he had been startled when the tunnel shook up and down—a motion he thought ominous, for normally when the tunnel shook it was from side to side. He immediately felt that something fundamental was wrong.
Both Lovell and Swigert thought that the bang—or shudder—had come from the lunar module, and as Haise emerged from the tunnel Swigert shot out of his seat and slammed the command-module hatch shut behind him. Haise scrambled to his seat—the right-hand one—for the master alarm was now sounding in his earphones. Swigert had noticed an amber caution light glowing overhead. It didn’t signal trouble in the oxygen tank, because that alarm system was still tied up by the low-pressure warning in the hydrogen tanks; rather, it signified trouble with the electrical system, the controls for which were near Haise. About this time, the Flight Surgeon, Dr. Willard R. Hawkins, noticed that the pulse readings for all three astronauts had shot up from about seventy to over a hundred and thirty.
The first disaster in space had occurred, and no one knew what had happened. On the ground, the flight controllers were not even sure that anything had. One reason for their ignorance was the imperfect nature of the telemetry from the spacecraft, which could not tell them directly that an oxygen tank had blown up. It could only report what the temperatures and pressures were in the tanks, whether certain voltages were within the proper limits, and whether certain equipment was on or off. This information had to be interpreted before the flight controllers could know what was going on, and the flight controllers were slow to make the correct interpretation, because, like everyone else at NASA, they felt secure in the knowledge that the spacecraft was as safe a machine for flying to the moon as it was possible to devise. Obviously, men would not be sent into space in anything less, and inasmuch as men were being sent into space, the pressure around NASA to have confidence in the spacecraft was enormous. Everyone placed particular faith in the spacecraft’s redundancy: there were two or more of almost everything. Even the flight controllers’ own training contributed to their confidence. For three months before the flight, they had flown the mission over and over in rehearsals called simulations. For these, a team of flight controllers took their places at the consoles in the Control Room while the astronauts got inside simulators—working models of the spacecraft, very much like the Link Flight Simulators that student airplane pilots use. Both groups were connected to computers that had been programmed to create problems likely to come up on the mission. The previous moon flights had gone so well that the flight controllers had complained on an earlier occasion to the men planning the simulations that these were too tough to be authentic. Accordingly, the simulations in preparation for Apollo 13 had dealt only with problems that were considered likely to arise; the controllers hadn’t wasted time on what one engineer called “four-point failures—way-out disasters.”
Before the flight controllers could admit the full scope of the present disaster, they went to great lengths to find explanations that would not involve a major failure of the spacecraft. It took them a quarter of an hour to get a rough idea of what had happened, and about an hour more to admit that the spacecraft was damaged beyond repair. At the outset, Liebergot, the EECOM, wasn’t particularly alarmed. Because he had happened to miss seeing the sudden rise in pressure in Oxygen Tank No. 2, it simply didn’t occur to him that the tank had blown out. There was such a cascade of problems that, not having noticed where they started, he didn’t know where to begin to look for their source. Since he had no reason to think in terms of the oxygen tank in the first place, he had to track the trouble backward step by step all the way through the electrical system. The only clue he had to start with was the electrical warning Haise had reported and a similar light flashing on his own console. When he tracked it down, he found that it signalled what he called “a Main Bus B undervolt.” A main bus is like a set of wall plugs. (Electricians also call it a distribution terminal board.) Electricity from the fuel cells—the generators—was fed into the buses, and then power was tapped out of them by the equipment that needed it. For redundancy, there were two main buses, A and B, and what Liebergot had found was that Main Bus B had suffered a significant drop in power, so that the equipment connected to it, which was half the equipment in the spacecraft, was in danger of failing. Up in the command module, Haise already had a sinking feeling, for, according to the mission rules, both buses had to be operating if the astronauts were to get the go-ahead to land on the moon.
Then there was a moment of relief. Haise saw the warning light above his head flicker out, and down in the Control Room the same thing happened on Liebergot’s console. Lovell reported to the CAPCOM that the power in the bus was back to normal. Over the loop, Liebergot suggested to Kranz that the trouble might not have been an undervolt at all but, rather, a problem with the instruments reporting the problem. In the next hour or so, they came back over and over again to this wishful explanation—what flight controllers call an “instrumentation failure.” Following this false trail, they told each other that perhaps everything was all right after all—though Haise now told the CAPCOM that “a pretty large bang” had been associated with the incident. Oddly, Kranz had not heard the astronauts mention a “bang” before. Now a light flickered on one of the panels on his console to indicate that one of the flight controllers—the INCO, who was in charge of the radios aboard the spacecraft—wanted to talk to him.
The INCO told Kranz about a communications “funny”—an aberration that doesn’t clear up immediately, as opposed to a “glitch,” which is a transitory one. At the time of the bang, the INCO reported, there had been an unexplained change in the width of the radio waves transmitted from the spacecraft: they had suddenly switched from a narrow beam to a wide one. Kranz was still not alarmed. The spacecraft radio was transmitting with the high-gain antenna—a sort of stick with reflectors that had to be aimed as precisely as a rifle—and it crossed Kranz’s mind that since the antenna ran on power from Main Bus B, the undervolt might somehow have caused the change; if that was so, then the funny should correct itself now that the undervolt had. Much later, it became apparent that when the side panel of the service module had been torn off and hurtled into space it hit the antenna, causing a change in the nature of. the radio signal.
Less than a minute had passed since the accident. A voice fro
m the spacecraft now said that the bang must have affected the gauge that reported the level of Oxygen Tank No. 2—first it had oscillated between twenty and sixty per cent, but now it was off-scale on the high side. This still did not cause Kranz or Liebergot to think that there might be a problem with the oxygen tank. They had been having trouble with the oxygen gauges all along, and they thought that the same trouble had cropped up once more. It was hard for anyone to get rid of the idea that the instruments were lying to them. Just then, Lovell reported that Main Bus B had no power in it at all and Main Bus A was beginning to show an undervolt, too; that is, one main bus had gone dead and the other was losing power. If both buses died, the command and service modules would be without any electrical power except a small amount available from three storage batteries to be used during the return through the earth’s atmosphere. Liebergot was confused. The two main buses, themselves paired for redundancy, were drawing their power from three redundant fuel cells; if one bus died, there was every reason for the other to hold up. While Liebergot pondered, there was a long silence, broken at last by Lovell, who asked, a little anxiously, “O.K., Houston, are you still reading Apollo 13?”
Lousma replied, “That’s affirmative; we’re reading you. We’re still trying to come up with some good ideas here for you.” Then, pressing a switch so that the astronauts couldn’t hear him, Lousma said hurriedly to Kranz, “Is there any kind of lead we can give them, or are we looking at instrumentation problems, or have we got real problems, or what?”
Thirteen: The Apollo Flight That Failed Page 2