What Would a Generation Ship Actually Need to Survive the Trip to Alpha Centauri?

<sub>Credit: NASA / Rick Guidice</sub>

*Written by OC Wanderer, author of Destiny Among the Stars, a litrpg sci-fi series set in the real stellar neighborhood. Published April 2026.*

**Short answer:** A generation ship bound for Alpha Centauri (4.37 light-years away) would need a founding population of at least 500 people to maintain genetic diversity, roughly half a square kilometer of onboard farmland, hundreds of metric tons of radiation shielding, and a working government capable of holding itself together across ten or more generations of people who never chose to be on the ship. The engineering problems are solvable in principle. The human problems might not be.

How Many People Do You Actually Need?

The minimum viable crew question turns out to have two very different answers depending on what you mean by "viable."

Frédéric Marin and Camille Beluffi at the University of Strasbourg built a Monte Carlo simulation called HERITAGE specifically to model multi-generational spaceflight population dynamics. Their 2018 result: a minimum crew of **98 people** can theoretically complete a voyage to Proxima Centauri b over 6,300 years, provided strict "social engineering" is enforced continuously, including yearly population audits, regulated breeding pairings, and hard constraints on reproduction. Without those controls, the crew goes extinct in roughly half of simulated runs.

Cameron M. Smith at Portland State University took a different approach in 2014. He modeled a shorter 150-year trip (roughly 5 generations) under Project Hyperion and asked a different question: not "can the crew survive?" but "will they arrive genetically healthy?" His answer was far grimmer. Smith found that populations below 500 people can lose up to **80% of their initial genetic diversity** within 30 generations. His recommended safe founding population: **14,000 to 44,000 people**, with 40,000 as the preferred conservative figure.

The gap between 98 and 40,000 reflects two genuinely different things being measured. Marin asks about mission survival with total social control. Smith asks about arriving with a gene pool capable of founding a colony. For a trip to Alpha Centauri at 1% of light speed (roughly 440 years, 14-18 generations), you are somewhere in between. You need enough people to absorb catastrophic accidents without going extinct and enough genetic diversity that the colonists who arrive can build a civilization, not a genetic bottleneck. Most serious proposals land between **500 and 2,500 people** as a practical working range: small enough to build and supply, large enough to survive.

A 2025 follow-up study from Marin's team added phenotypic modeling, tracking how genetic changes affect fertility, life expectancy, and miscarriage rates across centuries. Their revised finding: crews of approximately 500 individuals maintain allele frequency stability under normal shielding conditions. Under compromised radiation shielding, population genetics diverge dramatically.

**Bottom line:** The minimum safe crew is not a fixed number. It depends on your voyage length, your willingness to enforce population controls, and your definition of success. Five hundred people is probably the floor. Forty thousand is the ceiling if you want genuine long-term genetic security.

Closed-Loop Life Support: The Energy and Space Budget

![Tomato plants growing under red and blue LED lights in a NASA growth chamber](https://images-assets.nasa.gov/image/KSC-20180307-PH_CSH01_0015/KSC-20180307-PH_CSH01_0015~large.jpg) <sub>Credit: NASA / Cory Huston</sub>

A generation ship cannot carry 440 years of food. It must grow it. The only working model for a fully closed biological loop is ESA's MELiSSA project (Micro-Ecological Life Support System Alternative), which has been running since 1987 and represents the most rigorous engineering effort in this space.

MELiSSA's design uses five interlocked compartments. Thermophilic bacteria break down organic waste and inedible plant material, outputting carbon dioxide and ammonia. Photosynthetic bacteria consume the organic byproducts. Nitrifying bacteria convert ammonia to nitrates for use as plant nutrients. A photoautotrophic stage runs two parallel tracks: cyanobacteria (specifically Arthrospira platensis, also known as spirulina) for rapid oxygen production and high-protein food output, and higher plants (approximately 32 crops are under consideration) for carbohydrates, vitamins, and psychological variety. The crew is Compartment Five. The goal is total closure: every gram of waste feeds back into the cycle, and no consumables need to be resupplied from outside.

The space requirement for this system is substantial. Marin et al. (2019) calculated that a crew of 500 on a balanced omnivorous diet would require approximately **0.45 square kilometers of artificial agricultural land**. That is roughly 84 American football fields of growing area inside the ship.

The power requirement is harder to pin down. The most concrete real-world benchmark is Biosphere 2, the closed ecology experiment that housed 8 people from 1991 to 1993. It drew an average of **700 kilowatts**, peaking at 1,500 kilowatts, which works out to roughly 87 to 190 kilowatts per person. Scaling that to 500 people yields a rough range of 43 to 95 megawatts of continuous power just for life support. That is before propulsion, lighting, manufacturing, medical systems, or the several metric tons of computing infrastructure needed to manage the biological loops.

NASA's physiological data adds additional constraints. Without countermeasures, bone density drops at approximately **1% per month** in microgravity. Current ISS protocols require astronauts to exercise **two hours every day** simply to slow that loss. Multiply that by centuries, and you need not just gym equipment but a sophisticated ongoing medical capability capable of identifying and treating conditions that have never been seen in multi-generational spaceflight.

**Bottom line:** Closed-loop life support is theoretically achievable, but the system is enormous, power-hungry, and biologically complex. It must run continuously for centuries without catastrophic failure. MELiSSA's spirulina-and-plant design is the most developed practical model, but it has never been tested at ship scale or voyage duration.

Radiation: The Mass Problem That Could Sink Everything

Deep space is not empty. It is saturated with galactic cosmic rays (GCR): high-energy protons and heavy nuclei accelerated by ancient supernovae, moving at a significant fraction of light speed. On Mars, the annual radiation dose is approximately **672 millisieverts per year**, compared to 182.5 millisieverts per year on the ISS and 3 millisieverts at Earth's surface. A 440-year voyage would deliver cumulative doses thousands of times higher than any human has ever received.

The physics of shielding is unpleasant. A 500-day Mars mission (with no shielding improvements over current spacecraft) would require roughly **1,280 metric tons of isotropic passive shielding** just to meet permissible exposure limits. That figure comes from Barthel and Sarigul-Klijn's 2019 review in Progress in Aerospace Sciences, and it represents shielding mass alone, before propellant, structure, food systems, or crew.

Material choice matters enormously. Aluminum, the default spacecraft material, is actually one of the worst options for cosmic ray shielding because heavy nuclei hitting metal nuclei produce secondary neutron cascades that can be more damaging than the original particles. Hydrogen-rich materials including polyethylene, water, and synthetic fibers outperform metals significantly. A Geant4 Monte Carlo study by Gakis and Atri (2022) confirmed that composite low-atomic-number materials are the optimal passive choice. A layered design (outer hydrogen-rich layer to degrade the primary particles, inner neutron-absorbing layer) is the current best practice.

There is a hard limit on passive shielding, however. Oleg Semyonov's 2020 analysis showed that for spacecraft traveling faster than roughly **0.3% of light speed**, rarefied interstellar hydrogen gas transforms into an oncoming flux of relativistic nuclei that passive shielding cannot stop. The faster you go, the worse this gets. At genuinely relativistic speeds, a generation ship would need active electromagnetic shielding (a magnetic field strong enough to deflect charged particles around the hull, similar to Earth's magnetosphere). No such system has been built.

For a realistically paced generation ship traveling at less than 0.3% of light speed, passive hydrogen-rich shielding remains viable but extremely massive. For a faster ship trying to cut the 440-year trip to something closer to 50 years, active shielding becomes an engineering requirement, not an optional upgrade.

NASA's Human Research Program currently identifies radiation as one of the five core hazards of deep-space travel. Their own documentation acknowledges that there is "insufficient knowledge of the health effects of space radiation to provide recommendations on crew exposure limits for long-duration missions." That is a frank admission from the people responsible for protecting astronauts.

**Bottom line:** Radiation is not just a risk. It is potentially a civilization-ending problem for the mission. Hydrogen-rich passive shielding works at low velocities but adds crushing mass. Active electromagnetic shielding works at higher velocities but requires engineering that does not yet exist. The ship's speed determines which problem you have.

The Psychology of Centuries-Long Confinement

The longest isolation experiment ever run lasted 520 days. Six men lived in a simulated Mars mission habitat in Moscow, monitored by Basner et al. (2014) in a study published in PLOS ONE. The findings were instructive. One crewmember showed depression symptoms in **93% of mission weeks**. Two crewmembers generated **85% of all perceived conflicts**. Crew-to-ground-control conflicts outnumbered internal crew conflicts by five to one. Two crewmembers showed zero significant disturbance across the full 17 months.

That inter-individual variation is the dominant signal: identical conditions, wildly different outcomes. Now scale that to 500 people and 440 years.

Yin et al. (2023), publishing in Translational Psychiatry, synthesized the broader astronaut mental health literature. Reviews of astronaut mental-health data suggest that anxiety, depression, sleep disruption, and medication use are all significant concerns on long-duration missions, but the datasets are still small and heterogeneous. Arone et al. (2021) found that approximately 75% of astronauts used sleep medication, and 94% used some kind of medication during flight. For missions exceeding 600 days, the estimated severe mental health disorder incidence climbs above **60% annually**.

These numbers are from missions measured in months. A generation ship trip to Alpha Centauri at 1% light speed lasts 440 years. Every single person currently living on Earth who would board that ship would die onboard. Their children would die onboard. Their grandchildren and great-grandchildren would die onboard. The people who actually arrive at Alpha Centauri will be the fifteenth generation to have lived and died inside that hull, and they will have never seen a sky or felt a wind or understood the phrase "going outside" to mean anything other than a spacesuit.

The psychological literature on long-duration spaceflight has no data for this. The longest analog missions are 17 months. Everything about generation-scale psychology is extrapolation.

**Bottom line:** Psychological fitness for a 440-year voyage is a selection and institutional design problem, not just a medicine problem. The human mind was not designed for multi-generational confinement, and no study has run long enough to tell us how badly that will manifest at scale.

Governance Over Generations

![The European Parliament hemicycle in Strasbourg, showing tiered semicircular seating at scale](https://upload.wikimedia.org/wikipedia/commons/2/2c/European_Parliament_Strasbourg_Hemicycle_-_Diliff.jpg) <sub>Credit: David Iliff / Wikimedia Commons, CC BY-SA 3.0</sub>

The generation ship governance problem is one of the least-studied problems in the field, and arguably the most important. Every other system on the ship can fail and be repaired. If governance fails, everything else fails with it.

The foundational challenge is consent. The people who board the ship choose to go. Every subsequent generation did not. Children born on the ship inherit a mission they had no part in creating, a destination they cannot reach in their lifetimes, and a social structure built to serve a goal most of them will never see. Dion (2004), writing in Aviation, Space, and Environmental Medicine, identified this as the "common ingroup identity" problem: how do you create and maintain a shared mission identity in groups who did not voluntarily join the group?

Florian Neukart's 2024 analysis in arXiv identified the structural requirements: near-complete autonomy from Earth (signal delays make Earth-based governance physically impossible at interstellar distances), democratic governance principles redesigned for a closed population, legal frameworks balancing universal rights with ship-specific conditions, and explicit strategies to manage the cultural drift that will occur as the ship's population develops an identity distinct from Earth.

Resource equity is also a structural risk. On a ship with fixed resources and a mandatory social hierarchy (someone has to run the engines), stratification is not just possible, it is probably inevitable without active institutional design to prevent it.

**Bottom line:** The ship needs a government that can survive fifteen generations of people who never chose it, adapt to circumstances that cannot be predicted from Earth, and maintain mission coherence while also being humane enough that its citizens do not mutiny. No democracy has ever been designed for this. Most that have tried to create such systems across millennia have eventually collapsed or been overthrown.

What Would the Ship Actually Look Like?

If we take the 500-person lower bound and the engineering requirements derived above, the numbers look something like this:

- **Agricultural space:** 0.45 km<sup>2</sup> (roughly 84 American football fields of growing area, distributed across multiple decks) - **Power requirement:** 43 to 95 megawatts continuous for life support alone - **Shielding mass:** Hundreds to thousands of metric tons of hydrogen-rich material, depending on speed - **Exercise and medical facilities** capable of sustaining a full population across generations without external resupply - **Redundant life support systems**, because a single point of failure over 440 years is effectively certain to occur at some point

For the Smith 40,000-person figure, every number above scales by 80. The ship starts to look less like a vehicle and more like a sealed town with engines, with the political structure of a city and the social complexity of a city.

No such ship has been proposed with specific dimensions, but reasonable estimates for a 500-person vessel put it in the range of 1-3 kilometers in length. The ISS, for reference, is 109 meters long and houses 7 people.

**Bottom line:** A minimum-viable generation ship is not a spacecraft with a large crew. It is a city-scale enclosed ecosystem with a propulsion system attached, governed by an institution that has to outlast the parents, children, and grandchildren of everyone alive when it launches.

> [!lore] Love the science of the Centauri system? See it in action. *Destiny Among the Stars* is a LitRPG epic set in our actual stellar backyard. [Start Reading Today →](https://ocwanderer.com/storytime/story/destiny-among-the-stars)

Reality vs. Fiction

| What Popular Fiction Gets Wrong | What the Science Actually Says | | ------------------------------------------------------------------ | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | Generation ships carry hundreds or thousands of people comfortably | 500 is roughly the genetic floor. Fewer than that and you arrive at a destination with serious inbreeding risks, not a crew. | | The ship is basically a big submarine with more rooms | Supporting 500 people requires ~0.45 km<sup>2</sup> of farmland. The ship is an enclosed biosphere, not a vessel with cargo holds. | | You can build a generation ship with current-ish technology | Closed-loop life support at that scale has never been achieved on Earth. Biosphere 2 failed to maintain oxygen levels for 8 people. | | Radiation is a problem you solve with "shields" | At interstellar speeds, passive shielding adds hundreds to thousands of metric tons of mass. Above ~0.3% light speed, active EM shielding becomes necessary. That does not exist yet. | | The crew stays focused on the mission over generations | The people who arrive never chose to go. Multi-generational social cohesion in a sealed, resource-constrained environment has no empirical model. Partial analogs exist (isolated island societies, long-running religious orders), but none of them were sealed in a can. | | The captain is the problem | The governance structure is the problem. Individual authority is secondary to institutional design across generations. |

![Infographic Generation Ship](/api/uploads/galleries/originals/blog-assets/generation_ship_infographic.png "infograph")

How This Shows Up in *Destiny Among the Stars*

There are no generation ships in *Destiny Among the Stars*. Humanity has FTL drives, so the transit problem is mostly solved. The pressure moves to the colony itself: whether it can survive with limited supplies, thin institutional support, and long gaps between contact with the rest of humanity. Fast travel helps, but it does not remove the need for infrastructure, redundancy, and local self-reliance.

That is why the generation ship research still matters to the setting. A remote colony has to solve many of the same problems: food production, medical depth, replacement parts, social cohesion, chain of command, and population depth. If there are not enough people, or not enough genetic diversity inside that population, one bad failure can cascade through the entire settlement and stay with it for generations.

The minimum viable population question matters even more if the story ever uncovers a vault of stasis cells. A cache of sleeping colonists only helps if enough people are alive, healthy, and genetically diverse enough to support a stable population. Too few survivors, or too narrow a genetic mix, and the colony starts inside a bottleneck it may never get out of.

That fits the book better than the standard generation ship model. The tension is not a centuries-long voyage. It is what happens after arrival, when a settlement is far enough from the rest of humanity that weak institutions or missing infrastructure can quietly become fatal.

What Luca Thinks

*Luca Rossi is the twenty-year-old captain of the Triumph of Darron in* Destiny Among the Stars. *Hand him the research. This is what you get.*

Five hundred people minimum. Forty thousand if you want a real safety margin. That is a colony, not a crew.

The genetic side of it is what sticks with me. Marin's model works if you enforce breeding rules and keep auditing the population. Put that in plain language and it gets ugly fast. Somebody decides who gets to have kids, who they have them with, how many they get to have, and probably when. Then somebody audits the population every year to make sure everybody is still inside the plan. Maybe that keeps the mission alive. It still sounds like hell.

The power budget is not much kinder. Biosphere 2 puts the number at roughly 87 to 190 kilowatts per person. Multiply that by a few hundred people and you are basically running a small town with no outside grid and no room for a long failure. If power drops, food production drops with it, and then everything gets simple in the worst way.

What I keep circling back to, though, is the colony angle. In our universe we have FTL, so the trip is not the issue. The issue is what happens if somebody finds a vault full of stasis cells and starts thinking they just solved the population problem. Did they? How many are still alive? How many are healthy? How many are too closely related? How much real genetic variation is left once you count the survivors instead of the pods? If the mix is too narrow, the colony is already in trouble before anyone wakes up.

That part feels real to me. Not the giant ark ship. The moment when people are far from help, short on infrastructure, and one bad count away from learning the colony was doomed before it even started.

> [!lore] I didn't just research this for the article, I built a whole universe out of it. If you want a LitRPG where the sci-fi is as hard as the leveling system, check out *Destiny Among the Stars*. [Available now.](https://ocwanderer.com/storytime/story/destiny-among-the-stars)

Related Questions

**What is the minimum crew size for a generation ship?** It depends on what you are optimizing for. Marin and Beluffi (2018) found that 98 people can survive a 6,300-year voyage with strict enforced breeding protocols. Cameron Smith (2014) found that a crew below 14,000 risks losing up to 80% of its genetic diversity over 30 generations. Most practical proposals target 500 to 2,500 people as the working range for a trip of a few hundred years.

**How does radiation kill a generation ship mission?** Galactic cosmic rays deliver cumulative doses far above safe thresholds over a voyage lasting centuries. Passive shielding using hydrogen-rich materials helps at low velocities, but becomes impractically massive at higher speeds. Above roughly 0.3% of light speed, rarefied interstellar gas transforms into a relativistic particle flux that passive shielding cannot block, requiring active electromagnetic shielding that does not yet exist.

**Can you actually build a closed-loop life support system for a generation ship?** In principle, yes. ESA's MELiSSA project has been developing a five-compartment closed biological loop since 1987 using bacteria, cyanobacteria (spirulina), and approximately 32 higher crop species. In practice, the system has never been tested at generation-ship scale, and Biosphere 2 demonstrated that even Earth-based closed ecosystems are extremely difficult to stabilize. The engineering is achievable; the reliability across four centuries is the unresolved question.

**How long would a generation ship actually take to reach Alpha Centauri?** Alpha Centauri is approximately 4.37 light-years away. At 1% of light speed (a commonly used illustrative speed in generation-ship discussions), the trip takes roughly 437 years. At 5% of light speed, roughly 87 years. At 10%, roughly 44 years. Faster speeds require exponentially more propellant and create more severe radiation problems. Most generation ship proposals assume 0.1-1% of light speed.

**What happens to governance when the founding generation dies?** This is the central unsolved problem in generation ship sociology. The first generation chose the mission. Every subsequent generation inherits it. Maintaining institutional coherence across generations of people who did not consent to the voyage requires what organizational psychologists call "common ingroup identity," a shared sense of purpose that transcends individual experience. No institution in human history has been specifically designed for this challenge, though long-running religious orders, dynasties, and isolated island societies provide partial analogs.

**Would children born on a generation ship have legal rights?** This is an open question in international and space law. Children born in international space (beyond any nation's jurisdiction) do not fit neatly into existing legal frameworks. A generation ship operating with full autonomy from Earth would need to develop its own legal code, and that code would have to address fundamental rights questions including whether individuals can opt out of the mission, own property, refuse reproduction controls, or claim asylum if they oppose the ship's government.

Related Reading

- [How Long Would It Actually Take to Reach Proxima Centauri?](https://ocwanderer.com/blog/how-long-would-it-actually-take-to-reach-proxima-centauri) - [Could Proxima b Actually Be Habitable, or Is That Just Wishful Thinking?](https://ocwanderer.com/blog/could-proxima-b-actually-be-habitable-or-is-that-just-wishful-thinking) - [Why Is Alpha Centauri So Hard to Find Planets Around?](https://ocwanderer.com/blog/why-is-alpha-centauri-so-hard-to-find-planets-around) - [Does Time Dilation Mean Interstellar Travel Is Possible Within a Human Lifetime?](https://ocwanderer.com/blog/does-time-dilation-mean-interstellar-travel-is-possible-within-a-human-lifetime) - [How Far Is the Perseus Arm, and What Would It Take to Reach It?](https://ocwanderer.com/blog/how-far-is-the-perseus-arm-and-what-would-it-take-to-reach-it)

Sources

1. Smith, Cameron M. "Estimation of a genetically viable population for multigenerational interstellar voyaging: Review and data for project Hyperion." *Acta Astronautica*, vol. 97, pp. 16-29, 2014. DOI: 10.1016/j.actaastro.2013.12.013

2. Marin, Frédéric, and Camille Beluffi. "Computing the minimal crew for a multi-generational space travel towards Proxima Centauri b." *Journal of the British Interplanetary Society*, vol. 71, p. 45, 2018. arXiv: 1806.03856

3. Marin, Frédéric, Camille Beluffi-Marin, and Frédéric Fischer. "Genetic Evolution of a Multi-Generational Population in the Context of Interstellar Space Travels: Part II: Phenotypic Effects of Gene Expression." arXiv preprint, 2025. arXiv: 2502.07559

4. Marin, F., C. Beluffi, R. Taylor, and L. Grau. "Numerical constraints on the size of generation ships from total energy expenditure on board, annual food production and space farming techniques." *Journal of the British Interplanetary Society*, 2019. arXiv: 1901.09542

5. Barthel, Joseph, and Nesrin Sarigul-Klijn. "A review of radiation shielding needs and concepts for space voyages beyond Earth's magnetic influence." *Progress in Aerospace Sciences*, vol. 110, Article 100553, 2019. DOI: 10.1016/j.paerosci.2019.100553

6. Gakis, Dionysios, and Dimitra Atri. "Modeling the Effectiveness of Radiation Shielding Materials for Astronaut Protection on Mars." arXiv preprint, 2022 (revised 2024). arXiv: 2205.13786

7. Semyonov, Oleg G. "Radiation Conditions in Relativistic Interstellar Flight." arXiv preprint, 2020. arXiv: 2004.10079

8. Yin, Yishu, et al. "Long-term spaceflight composite stress induces depression and cognitive impairment in astronauts: insights from neuroplasticity." *Translational Psychiatry*, 2023. DOI: 10.1038/s41398-023-02638-5. PMC: 10632511

9. Basner, Mathias, et al. "Psychological and Behavioral Changes during Confinement in a 520-Day Simulated Interplanetary Mission to Mars." *PLOS ONE*, 2014. DOI: 10.1371/journal.pone.0093298

10. Arone, A., et al. "The Burden of Space Exploration on the Mental Health of Astronauts: A Narrative Review." *Clinical Neuropsychiatry*, 2021. DOI: 10.36131/cnfioritieditore20210502. PMC: 8696290

11. Neukart, Florian. "Toward the Stars: Technological, Ethical, and Sociopolitical Dimensions of Interstellar Exploration." arXiv preprint, 2024. arXiv: 2402.15536

12. Dion, Kenneth L. "Interpersonal and group processes in long-term spaceflight crews: perspectives from social and organizational psychology." *Aviation, Space, and Environmental Medicine*, 2004. PMID: 15267074

13. ESA MELiSSA Project. "Micro-Ecological Life Support System Alternative." European Space Agency. https://www.esa.int/Enabling_Support/Space_Engineering_Technology/Melissa

14. NASA Human Research Program. "Space Radiation Element." NASA. https://www.nasa.gov/hrp/elements/space-radiation/

15. NASA. "Counteracting Bone and Muscle Loss in Microgravity." NASA. https://www.nasa.gov/humans-in-space/living-in-space/counteracting-bone-and-muscle-loss-in-microgravity/