40 crew-days: why Vast's first space station skips closed-loop life support

Forty crew-days. That is the total human presence Vast has booked across the entire orbital life of Haven-1, the single-module station it plans to launch in the first quarter of 2027 — four astronauts, roughly ten days each, four missions.

Forty crew-days is also about one tenth of the number NASA's own life-support engineers used when they worked out where closing the loop on air and water starts paying for itself. The station the commercial sector expects to inherit the ISS's recycling systems is buying Shuttle-era consumables instead. The cost curve says that is the right call.

The break-even curve NASA already published

The most useful document in this entire field is fourteen years old and almost nobody in the commercial station business cites it. In late 2011 the Johnson Space Center life-support team was given a specific job: define requirements, build a system concept and produce a preliminary design for a non-planetary habitat called the Deep Space Habitat. The mission was crew of four for 388 days, departing low Earth orbit for the asteroid 2008 EV5.

They had some freedom in how the crew got there. The habitat would launch unmanned, mate with a Space Exploration Vehicle, and loiter in high Earth orbit for up to 825 days while the crew travelled separately, arriving aboard a Multi-Purpose Crew Vehicle. That unmanned segment is where most of the complexity lives, and it matters later.

The core architecture was pulled from hardware already flying on the International Space Station, plus a handful of technologies sitting at moderate readiness on the ground. What the team did with the resulting design was unusual for NASA. Instead of assuming the most capable system was the right one, they ran a trade study across mission durations from 100 days to 900 days, holding the habitat's hard shell and its power architecture — solar photovoltaics with fuel-cell storage — fixed while varying the part everyone argues about.

They compared open, partially closed and fully closed architectures side by side. The metrics were overall recovery efficiency in the water system and how efficiently carbon dioxide got consumed and fed back into the air subsystem. They ran cases with and without laundry, and worked through a range of oxygen generation and carbon dioxide reduction options. Then they applied a volume cost equivalency, because a closed system is mostly about the volume of hardware it occupies, and evaluated the return on investment for each architecture.

Two results came out of that sweep, and they are the reason this piece exists.

The first: break-even points for the open loop and partially closed loop architectures lie close to the 2008 EV5 mission duration. Roughly 388 crew-days is where the arithmetic tips.

The second: equivalent system mass for the closed loop architectures was considerably higher than for the open or partially closed systems. Closing the loop is not free even when it pays. It adds mass, and mass is the thing you pay launch prices for.

So the team went back to the requirements and cut scope. The final architecture concept contains only partially-closed air and water systems, because the breakeven point for some of the closure technologies was not achieved with the mission duration. That is a NASA engineering team, on a flagship deep-space mission, choosing not to close the loop — and the reason given in the paper is duration, not capability.

Haven-1 books about 40 crew-days across its entire life. On that curve it sits at roughly one tenth of the point where the trade flips.

What open loop means when you actually build it

BBC Science Focus reported in June 2025 that Haven-1 runs an open-loop life support design borrowed from earlier NASA technology, of the kind used on the Space Shuttle. The practical version of that statement: the air and the water arrive with the module, and the used air and used water leave with it. Nothing on board converts waste back into consumables.

The hardware around that decision is modest. Vast lists 45 m³ of habitable volume, 80 m³ pressurized, a 14,600 kg dry mass and a 13.2 kW deployable array, at 425 km and 51.6° inclination. BBC Science Focus put the interior at roughly the volume of a small tour bus. Consumables ride up on the shelves — food, water and other life-support cargo — and they come down with the crew vehicle at the end of a visit.

The build sequence is the interesting part, because it shows how seriously the consumable path is being engineered. Vast announced on 20 January 2026 that integration had begun, starting with the pressurized fluid systems: thermal control, life support and propulsion tubes, plus component trays and tanks. All of it goes through pressure, leak and functional testing, then a round of system environmental tests at NASA's Neil Armstrong Test Facility later in 2026. Launch moved to the first quarter of 2027.

Even an open-loop station buys closed-loop-adjacent engineering. NASA tested Vast's trace contaminant control system in an environmental chamber at Marshall Space Flight Center under a reimbursable Space Act Agreement — the same chamber used to test the ISS life-support system. Filters that scrub hazardous chemicals off the cabin air are not the expensive part of regenerative life support, but they are the part a crew of four cannot skip.

Five programmes, five answers

Habitat volumes and crew models as published by each operator; the ISS row is included for life-support reference rather than volume comparability. Sources: Vast, Axiom Space, NASA, ESA.

The table is not a league table. Volume and duration are the two inputs that move the break-even calculation, and each programme is buying a different point on that curve. Axiom specified complete environmental control and life support for Hab-2 because it expects to berth eight people. Starlab's own NASA documentation describes recovering purified water from urine to cut resupply. Vast has said almost nothing about Haven-2's architecture, which is consistent with a company that intends to keep moving along the curve as its missions get longer.

The ISS is the outlier

It is worth being precise about what the space station actually does, because the recovery numbers get quoted out of context constantly. NASA's Water Recovery System reclaims wastewater including urine, cabin humidity condensate, and water from spacesuit recharge, then pushes it through multi-filtration beds and a catalytic oxidiser. Purity is checked by conductivity sensors, and anything that fails gets reprocessed. That system recovers about 90 percent of the water on station.¹

The oxygen loop is a closed machine. The oxygen generation assembly electrolyses recovered water into oxygen and hydrogen. Oxygen goes to the cabin; hydrogen is either vented or handed to the carbon dioxide reduction assembly, which runs it with exhaled CO₂ through a Sabatier reactor. Methane goes out. Water comes back.

Europe's Advanced Closed Loop System extends the same idea one step further by cracking half the station's CO₂ directly back into oxygen, sized for three astronauts and installed as one two-metre rack. It saves roughly 400 litres of water per year. A single rack.

None of that is wasted engineering. It is the accumulated cost of more than 25 years of continuous crew. Four astronauts and about a gallon of water each per day, covering drinking, food preparation and hygiene, works out to roughly 144,000 litres through the station's pipes since 2000. At 90 percent recovery, something like 14,000 litres still had to be launched.

¹ NASA's Water Recovery System recovers about 90 percent of station water system-wide. A separate 70–85 percent figure in NASA's 2012 habitat study refers only to the urine distillation subsystem, one component of that system. The two numbers measure different things, and they get conflated constantly.

Which is the entire argument, and it is an argument about duration. NASA's own Next Generation Life Support programme puts it plainly: the state-of-the-art systems on the ISS are only partially closed and still require frequent resupply. They exist anyway, because the station has been crewed long enough for the fixed cost to amortise. Take the crew-days away and the same logic points the other way.

Launch cost keeps moving the line

Every number above is a function of one variable: what a kilogram costs to put in low Earth orbit. Bringing water up to the station runs into the tens of thousands of dollars per resupply. Researchers writing in Eos, the American Geophysical Union's science magazine, flagged the other half of the trade — the current recycling systems are energy-intensive enough that they may not scale cleanly to longer missions, so closing the loop buys logistics savings and spends power.

NASA's 2012 study priced the other end of that trade in equivalent system mass. It also had to launch about 450 kg of water purely to replenish spacecraft consumables, with the product water tank filled on the ground holding about three days of supply as a buffer, around 85 kg. Mass you must launch to carry a mass you then recycle is the whole argument against closing the loop before it pays.

As launch costs fall, both sides of that ledger move. The fixed mass of a closed loop becomes cheaper to carry, and the per-kilogram cost of the consumables it replaces becomes more expensive to skip. The break-even point moves left — closer to the short-mission end — with every reuse success. A station designed for 40 crew-days in 2027 and 4,000 crew-days in 2035 is not contradicting itself. It is reading the same curve at two different points on the horizontal axis.

Vast's own roadmap is the cleanest evidence that this is how the industry thinks. Haven-1 launches first as a single module with four short missions. Haven-2 adds a seven-metre core module and four more modules between 2030 and 2032, targeting continuous presence after the ISS is decommissioned. Same company, same engineering, opposite end of the life-support curve.

The trade looks like a step backwards only if you treat closed-loop life support as a technology ladder where every rung is a win. It is not a ladder. It is a curve with a crossing point, and NASA mapped the crossing point in 2012. The station that buys consumables because it will fly four short missions is reading the curve. So is the station that recycles because it will host eight people for years. Different missions, same arithmetic.