Read Seveneves: A Novel Online
Authors: Neal Stephenson
“Fuel flea?” Jiro had used the term before. No one else knew what it meant. It had gone in one ear and out the other, just another bit of the tech jargon that was so ubiquitous on Izzy. Now that fuel fleas were killing people, it was time to learn about them.
“A tiny piece of uranium or plutonium that has gotten loose from a ruptured rod. As it throws off alpha particles, it zigs and zags
around the room—conservation of momentum. So it hops around like a flea. The point is, it is small and it makes a lot of alpha. It lodged in a diverticulum in his bowel. It burned through his bowel wall and started a bleed that could not stop.”
Everyone pushed back their food.
“Okay,” Markus said. “We eat in
New Caird
.”
Once they had finished their meal, Markus told everyone that they needed to sleep, since they had a busy few days ahead of them. Jiro volunteered to take the first watch, and so the rest of them slept while Jiro stayed up going through logs and notebooks, assembling a picture of all that had happened on
Ymir
’s journey.
Suddenly they had a lot of space to spread out in. Dinah was tempted to retreat to the far end of the
New Caird
and get some privacy, but Markus insisted that everyone sleep down in the command module.
New Caird
might be free of radioactive contamination, but it was exposed to the direct hazards of space. A bolide strike would kill anyone in it. Whereas a beta-emitting particle, inhaled into the lung, would take days or weeks to incapacitate the victim—time during which they could do useful work.
So Vyacheslav ended up sleeping in one of the berths on
Ymir
’s crew accommodation level, while Dinah and Markus shared another. Somewhat to her surprise, they actually managed to have sex, a thing that had occurred only once since the White Sky. It was a sly surprise, not the athletic banging around that they had enjoyed the first few times they had done it, back in the good old days when the Hard Rain had seemed far in the future and the Cloud Ark had still felt like an isolated research colony.
Ymir,
now separated from the rest of the human race by millions of kilometers’ distance and several thousand meters per second of delta vee, now had some of that old feel to it. And despite the ghoulish scene that had greeted them on arrival, Dinah liked it here—it was the space equivalent of one of Rufus’s old mining camps—and didn’t really want to go back.
But they were supposed to be saving the human race, not enjoying
an exotic holiday, so she tried to get some sleep. When Markus’s alarm went off five hours later she peeled out of the bag they’d been sleeping in and did her best to clean up and get into some fresh clothes.
Ymir
had long ago turned into a smelly bachelor pad, short on toiletries, and, as they discovered while rooting around in the common area, on food. Sean had definitely been killed by the fuel flea in his gut, but he had likely been weakened before then by malnutrition, and even by lack of oxygen. For the systems that the crew had been using to replenish
Ymir
’s air supply were not in the best condition. The new arrivals were awakened twice during their sleep cycle by alarms from the life support system, which Jiro silenced and dealt with.
When they were all awake, they ate food from the stores they had brought with them and listened to a briefing from Jiro.
“Let me tell you what happened to this expedition,” he said. And then he told them the story as he had pieced it together from the logs left behind by the dead.
The failure of the radio, shortly after the beginning of the mission, had been caused by a defective part for which there was no replacement: a simple, stupid oversight. The longest leg of the trip—the year and a half spent coasting from the L1 gate to Grigg-Skjellerup—had consisted of lengthy stretches of boredom interrupted by occasional panics, most of which had to do with the life support system. This was based on using sunlight to grow algae, a process that worked well in the lab but had turned out to be difficult to sustain on
Ymir
. The newest arklets in the Cloud Ark had benefited, in this respect, from lessons learned operating such systems in the time since Zero, but
Ymir
had been built and launched very early, using systems that now seemed painfully out of date.
Once they had reached “Greg’s Skeleton” and thereby gotten access to vast amounts of water, they’d been able to make oxygen by splitting H
2
O, and life had improved. Until then, however, they’d been oxygen hungry and tense, trying to keep their consumption of air and food to a minimum by floating listlessly in their sacks watching
the same DVDs over and over again. Health, and mental status, had suffered.
They broke the shard from Grigg-Skjellerup using small mining charges planted by hand, or by robots programmed by Larz. Into its nose they embedded the command module, making themselves comparatively safe from cosmic radiation and bolides for the first time since the beginning of the mission. Life began to improve. They started excavating the access tunnel into the core. Into the aft end of the shard they inserted the reactor system, letting it melt its way into the ice. Around it, in the heart of the shard, they began to excavate a cavity and sculpt out hoppers: containers designed to hold broken-up ice produced by the mining robots. Twelve augers—long, spiraling ice movers, like the ones used to transport grain into elevators—were set up to convey that loose ice from the hoppers into the space surrounding the warm reactor vessel, where it would melt and be pumped into the core itself. Meanwhile, a separate corps of robots worked on the outside of the shard, melting the ice a little bit at a time, mixing it with the fibrous material they’d brought with them, and letting it refreeze into the much tougher material known as pykrete.
The “steampunk” propulsion system had basically worked as planned—though not without a lot of tinkering and head scratching—on the first “burn” that had put it on the course back to L1. There had, however, been some problems with the augers that were used to feed ice into the reactor chamber. The augers received their inputs of ice from hoppers that had to be filled up by “mining” solid ice from the inside of the shard, a process for which robots were well suited, and so nothing worked at all without the assistance of a small army of robots conveying flakes of ice from mine head to hopper like ants dismantling a loaf of sugar. This part of it had actually worked. But some of the pieces of ice being mined by the robots had little rocks in them. These jammed the augers. Jams could often be repaired by operating the auger in reverse for a short time, but sometimes
a robot, or even a person in a space suit, had to be sent to pry a rock out of the mechanism. An auger accident had led to the death of one member of the crew.
During the months between that first burn and their arrival at L1, Larz did some programming work on the robots, trying to teach them not to collect rocky ice. They conducted a number of system tests intended to make sure that the problems they’d experienced the first time around wouldn’t be repeated during the critical second burn. These ranged from small-scale tests on individual robots all the way up to full dress rehearsals where the entire system would be energized and the reactor turned on to generate thrust for a few minutes.
It had been during the first of those dress rehearsals when something had gone wrong in the core, resulting in damage to the jacket of a fuel rod.
Jiro had an idea as to what had gone wrong.
Ymir
’s reactor used water—the melted ice of the comet core—as its moderator. In nuclear engineering, that meant a medium that slowed down the neutrons hurled out by fission reactions, making them more likely to stick around long enough to trigger more such reactions. In the absence of an effective moderator, the neutrons would mostly escape from the system without doing anything useful.
Between being as dead as a doornail and running out of control was a narrow band of normal and healthy power output in which basically all commercial reactor operations happened. The essential problem with
Ymir
’s reactor was that its moderator—being a naturally occurring substance—was impure and unpredictable. The water that flooded into the chamber for the first dress rehearsal had been melted from ice a few months earlier, around the time of the initial “burn,” and had been sitting in the plumbing system ever since then. There, it had been in contact with rocks and grit that had made it through the augers. It had leached various minerals out of that rock, and become something other than pure water. When the reactor
was started and the pumps turned on, that impure water was drawn through screens and filters intended to exclude all the debris. But it was nonetheless impure water, and when introduced to the core, it failed to perform its function as a moderator. The reactor was sluggish to get going. With the advantage of hindsight, it could be seen that its neutron economy was suppressed, poisoned by the impurities in the water. Overreacting to the slow start, the operators had pulled the control blades out farther than they would have otherwise. But once the first rush of impure water had been flushed through the system and blown out the nozzle, it had been replaced by relatively pure water, only just now melted from the ice. The reactor’s power had surged, producing a sudden buildup of fission products inside the fuel rods. Some of those would have been gases such as krypton and argon. The gases would have created pressure. Fuel rods were engineered to withstand it, but one of them had failed and ruptured. Possibly it had left the factory in excellent condition but been damaged en route by a nanometeoroid that had left a microscopic flaw. In any case, for whatever reason, the rod burst open and began to spill out the highly radioactive “daughters” of nuclear fission, which had become mixed with the steam being blasted out the rocket’s nozzle.
Most of the fallout had, therefore, dissipated into space. But the whole point of a rocket nozzle was to convert the thermal energy in the gas—its heat—into velocity. The faster the steam went, the colder it got, until the steam near the nozzle exit was so cold that it actually began to condense into snow. Tiny particles of fallout made excellent nuclei around which a snowflake might begin to form. Some of that snow had stuck to the ice walls of the nozzle bell.
The most likely explanation for what had happened next was that one of the robots crawling around in that area maintaining the shape of the nozzle had become contaminated with a mixture of alpha-emitting fuel fleas and beta-emitting daughters, and tracked the material to a location where it had been transferred to the glove of a space suit—possibly by a mechanism as simple as a spacewalker
reaching down to brush some ice from the claw of a Grabb, or planting a foot in a location where a contaminated Grabb had stepped. The contamination had then been brought into the command module when the spacewalker came indoors. They might not even have known about the burst fuel rod, so they might not have been checking for contamination. Or, as suggested by Sean’s note, their Geiger counters might have broken down, one by one, rendering them blind to the presence of radioactivity in their environment. In any case, the particles had spread around the command module. Some men had inhaled them, some had swallowed them. They hadn’t been healthy to begin with.
IN ANY CASE, THE GOOD NEWS, IF IT COULD BE SO CALLED, WAS THAT
the reactor and the engine basically worked. The improvements Larz had made to the robots’ mining programs had led to fewer rocks in the hoppers, and fewer jammed augers, during the L1 burn. Since then, Nats had been crawling around in the hoppers identifying rocks that had sneaked in anyway, and pushing them away from the augers. The damage to the fuel rod would have been a major catastrophe by Old Earth standards—had it happened on an earthbound reactor. Here, it was messy, and had already been fatal to a few. But everything still worked. Yes, the
New Caird
expedition would be bringing a radioactive disaster right into the middle of the Cloud Ark, but once they drew close enough they would jettison the reactor and let it fall into the atmosphere.
Forty-eight hours, give or take a few minutes, now remained before Earth would loom huge below them, and the nadir surface of the shard would sweat and steam as the radiant heat shining up from the incandescent air softened, melted, and vaporized the ice. It was then that they would have to pull out the control blades and execute
Ymir
’s next big burn. First they would have to spin the whole ship around so that she was flying “backward,” her nozzle bell pointed in
the direction of movement. For the delta vee they needed was a negative one—a braking, as opposed to accelerating, burn.
For spin moves, all spaceships were equipped with thrusters, not powerful enough to impart big delta vees but capable of rotating the ship as a whole into the desired attitude so that the main engine was pointing in the right direction. As a rule the thrusters were more effective when they were situated out toward the “corners” of the vehicle, where they could exert more leverage and crowbar the thing around with minimal thrust. Not knowing what they were going to find at Grigg-Skjellerup, the mission planners for
Ymir
had packed aboard a collection of modular thruster assemblies that basically consisted of little rocket engines, propellant tanks, wireless control links, and hardware for anchoring them into ice. A cursory survey of
Ymir
and a look at the dead crew’s records made it clear that Sean and his crew had embedded those packages into the ice at suitable locations: one complex up at the nose with nozzles aiming in four perpendicular directions, and four more spaced around the fattest part of the shard.
Now that
New Caird
was docked, her engine could also be put to use in getting
Ymir
spun around. But this one maneuver—a 180-degree flip, which would have seemed comparatively simple in a small craft such as an arklet—was fraught with difficulties and complications in something as huge and asymmetrical as
Ymir
. Anticipating a need to use the thrusters, Dinah sent robots out to inspect them during that first “morning,” and Vyacheslav suited up and went out to do a bit of troubleshooting on a propellant line that had somehow become kinked. But so ponderous were the shard’s movements that the actual rotation, end-over-end, consumed eight hours, and tweaking it into precisely the right orientation then took another six.