The Dream and the Void

The observation dome holds the world she left, suspended in absolute black, and Anya Sorokina is the only human being who can see it from this angle without a visor between her eyes and the vacuum. She floats with one hand resting on a handhold bolted to the dome's equatorial ring, her boots magnetically coupled to the floor plate, her body a small warmth in a cathedral of aluminum and glass and radiation-hardened polymer. Below her, where there should be a floor of rock and dust, there is only the curved interior of the dome itself, the structural ribs of it converging at her feet, and beyond that, through the triple-pane viewport, the silver-white limb of Shackleton Crater rolling in slow shadow toward the south pole of the Moon. Above the crater rim, in the same unblinking black, hangs the Earth. It is half in day and half in night, terminator sharp, a sliver of ocean glinting where the sun has not yet reached it. She has been at the Lunar Gateway for eleven months, four days, and an unknown number of hours that she has stopped counting because the station's master clock, tied to a Cesium standard at the White Sands ground station, has slipped by 1.4 seconds relative to UTC and she cannot decide whether the discrepancy is comforting or terrifying.

She is not lonely in the way the early astronauts described. She has a video uplink to a husband in Houston, a sister in Vladivostok, a six-year-old nephew who sends her crayon drawings of the Moon rendered in slightly the wrong shade of yellow. She has the slow, asynchronous messages of her ground support team, the weekly psychological check-ins with a flight psychiatrist who is very good at her job and not very good at pretending that an eight-second round-trip latency does not complicate the meaning of "real-time." She has, more importantly, the station. The Gateway is a strange object, half-finished by the standards of its designers, embarrassingly complete by the standards of every other structure ever built beyond low Earth orbit. Its modules were launched on three separate heavy-lift vehicles and assembled in a near-rectilinear halo orbit that keeps it continuously above the lunar south pole, never setting, never eclipsed, a relay point between the surface and the deep-space tracking network. It houses her and, for the moment, no one else.

The fact of being alone, while novel, is not what makes her stare. What makes her stare is the realization, arriving with the slow, mechanical certainty of a clock's second hand, that she is the last line of defense for every system that keeps her alive. There is no other human being aboard who can read the pressure decay in the oxygen partial-pressure log and recognize the signature of a slowly leaking heat-exchanger coil. There is no one to back her up when the CO₂ scrubber's primary sorbent bed reaches saturation at 03:00 station time and the backup bed, which should have been regenerated during yesterday's maintenance window, was not, because yesterday's maintenance window contained a two-hour video call she did not want to cut short. There is no one to take over if the centrifuge in the medical bay, the one that is supposed to provide her with 1 g for two hours each day to keep her bones from surrendering their calcium at a rate of one to one and a half percent per month, throws a bearing fault and she has to fix it with the limited spares aboard. There is no one to share, in the precise and technical sense of the word, the air she breathes.

This is the condition. It is the condition of being the first person to live permanently in space outside low Earth orbit, and it is, she understands, a condition that cannot be solved by adding more people. Adding more people adds more systems, more failure modes, more kilograms of oxygen to be supplied, more kilograms of carbon dioxide to be removed, more kilograms of water to be recycled, more kilograms of food to be stored or grown, more kilograms of feces to be processed, more kilograms of psychological weight to be carried between conversations. Adding more people, in the language of the engineers who designed her station, is called *consumables loading*. The phrase is a euphemism in the same way that *velocity at impact* is a euphemism. The people who join her will be additional mass, additional heat, additional entropy, additional risk. They will not, by their presence, make her more alive. They will only make the system that keeps her alive more complex, and a more complex system is, by every law of reliability engineering, a less reliable one.

The first permanent residents of space will live inside a system whose physics does not care whether they are guests or inhabitants. The system will fail on its own schedule. The system will not be resupplied within a meaningful timeframe once a mission is sufficiently far from Earth. The system will, in the end, kill them if they do not understand it well enough, build it carefully enough, and tend it patiently enough. This book is about that system. It is about the engineering of closed worlds.

1.1 The Long Arc of Staying: From Salyut to a Permanent Place

The history of human spaceflight, when told as a sequence of "firsts," reads like a confident march. Yuri Gagarin, first human in space, 1961. Valentina Tereshkova, first woman in space, 1963. Alexei Leonov, first spacewalk, 1965. Neil Armstrong, first human on another world, 1969. The cadence is the cadence of a culture that has learned, in the span of a single decade, to leave its planet. What the cadence does not show is that every one of these milestones was achieved by a *visitor*. The first human in space spent 108 minutes in orbit. The first human on the Moon spent 21 hours and 36 minutes on the surface. Even the longest of the Apollo missions, Apollo 17, kept its crew on the lunar surface for a little over three days, and even the longest of the early space station missions, the Salyut series, kept its cosmonauts aloft for a few months before returning them home.

The Salyut program is, in retrospect, the first laboratory in which the engineering problem of *staying* was confronted honestly. Salyut 1, launched in 1971, hosted three different crews. The first crew, aboard Soyuz 10, failed to dock properly and never entered the station. The second crew, Georgy Dobrovolsky, Vladislav Volkov, and Viktor Patsayev, spent 23 days aboard and died on reentry when the Soyuz 11 capsule depressurized at altitude. They were the first humans to die in space, and the investigation that followed forced the Soviet space program to confront a truth that the early engineering culture had been reluctant to face: that a habitable spacecraft was not a vehicle but a *building*, and buildings are not finished when they are launched. Buildings settle. Buildings leak. Buildings require continuous, physical, sometimes grimy maintenance from people who know where the pipes run and where the spare parts are stored. The Soyuz 11 failure was a *vehicle* failure, not a *habitat* failure — the station itself performed nominally through the entire 23-day mission. But the accident changed the philosophy of Soviet space medicine, which had been willing to accept mission durations of a few weeks as a reasonable extension of the aviation model of flight. After Soyuz 11, the program began to treat the human body and the human habitat as a single, coupled system whose endurance had to be tested as a whole.

Salyut 4, 5, 6, and 7 progressively extended the duration of single missions. Salyut 6, launched in 1977, was the first to support crew rotations and resupply missions with the Progress cargo vehicle, and it is in Salyut 6 that the architecture of long-duration spaceflight was effectively defined. Cosmonaut Vladimir Kovalyonok and flight engineer Valery Ryumin spent 139 days aboard Salyut 6 in 1978, demonstrating that humans could, in principle, live in orbit for a substantial fraction of a year. Salyut 7, launched in 1982, hosted a cr