How Astronauts Recycle Water In Space

On the International Space Station, much of the water an astronaut drinks was recently urine, sweat, or moisture exhaled into the cabin air. NASA's Environmental Control and Life Support System (ECLSS) now recovers about 98 percent of the water aboard, using distillation, filtration, and chemical treatment. The 98 percent figure was announced in 2023, after a Brine Processor Assembly joined a system that had been running since 2008. It matters because the same closed-loop plumbing is the template for any crew that has to survive far from a resupply rocket.

Why recycling water is a launch-cost problem, not a hygiene problem

Every kilogram sent to low Earth orbit costs a great deal of money in propellant, so shipping drinking water to the station the way a warehouse restocks bottled water is not a realistic long-term plan. A crew member needs about one gallon (roughly 3.8 liters) of water a day for drinking, food preparation, and hygiene, according to NASA engineers quoted in Environmental Science & Engineering Magazine. Multiply that by a crew of seven, which was the average size aboard in 2021 through 2023, and the station needs on the order of 27 liters of water every day. Over a year, resupplying all of that from the ground would add up to roughly 10,000 liters, or ten tonnes of water, before counting the tanks and packaging that carry it.

That is the real reason the ECLSS exists. It is not a comfort feature or a curiosity for schoolchildren asking whether astronauts drink their own urine. It is a mass-budget solution to a physics problem: rockets can only lift so much, and every kilogram spent on water is a kilogram not spent on food, spare parts, or experiments. NASA water subsystem manager Jill Williamson has made the same point in plain terms, noting that taking less water and oxygen into space leaves room for more science on the launch vehicle.

NASA's answer was to treat the station less like a vehicle stocked from Earth and more like a small closed municipal water utility. Instead of hauling water up and venting wastewater overboard, the system captures the water that would otherwise be lost: humidity condensed from crew breath and sweat, and water reclaimed from urine. A companion article on the station's life support systems covers the air side of the same loop. This article stays with the water.

NASA's engineers describe the loop in one line. Christopher Brown of Johnson Space Center put it this way: "You lose two pounds of that and the other 98% just keeps going around and around." Before the Brine Processor Assembly, the system topped out at a recovery of about 94 percent.

A useful comparison is a modern building's air-conditioning system. Rather than venting conditioned air outdoors and pulling in fresh, unconditioned air to replace it, the building recirculates and reconditions what it already has, because reconditioning is cheaper than replacing. Water on the station works the same way, with one important difference: the "reconditioning" step includes a spinning still that evaporates urine at low pressure.

The architecture: two racks, two assemblies, one loop

The ECLSS water hardware on the U.S. side of the station is called the Water Recovery System (WRS). It has two main parts, the Urine Processor Assembly (UPA) and the Water Processor Assembly (WPA), housed in two refrigerator-sized International Standard Payload Racks in the Node 3 module. NASA delivered the hardware on November 14, 2008, according to a 2024 status paper by NASA Marshall engineers that this article relies on heavily.

Three streams of wastewater feed the system. Crew urine is collected in the Waste and Hygiene Compartment and chemically pretreated. Humidity condensate is pulled out of the cabin air by condensing heat exchangers in the air-conditioning loop. When the Sabatier reactor is working, product water from carbon dioxide reduction adds a third stream. According to the 2024 paper, that reactor had not been in service since October 2017 because of a hardware failure, with replacement hardware planned.

The output goes to a potable water bus held at roughly 230 to 280 kilopascals (19 to 26.5 pounds per square inch gauge) so water is available on demand. Its users include the crew's drinking dispenser, the hygiene and food systems, the urinal flush water, spacewalk suit coolers, scientific payloads, and the Oxygen Generation Assembly, which splits water into hydrogen and oxygen by electrolysis. The same paper says a reserve of at least 818 liters of potable water is stored aboard to cover contingencies.

Stage	Hardware	Input	Output
Collection and pretreatment	Waste and Hygiene Compartment, chemical pretreatment	Urine and flush water	Stabilized urine
Distillation	Urine Processor Assembly (Distillation Assembly)	Pretreated urine	Distillate and brine
Brine recovery	Brine Processor Assembly	Brine	Water vapor and solid brine residue
Filtration	Water Processor Assembly filters and beds	Distillate plus humidity condensate	Filtered water
Oxidation	Catalytic reactor	Filtered water	Water free of trace organics
Polishing and biocide	Ion exchange bed	Reactor output	Potable water with iodine
Distribution	Potable water bus and storage tanks	Potable water	Drinking, hygiene, electrolysis, payloads

Table: The U.S. segment water loop, simplified from the NASA Marshall and Johnson Space Center papers cited in this article.

Inside the Urine Processor Assembly: a still that spins

On Earth, distillation leans on gravity: liquid stays at the bottom of the pot, vapor rises, and the two separate on their own. In microgravity there is no "down," so liquid and vapor would mix into foam. The Urine Processor Assembly gets around this by using rotation as a substitute for gravity.

Its core is the Distillation Assembly, a rotating centrifuge. According to the 2024 NASA Marshall status paper, the Distillation Assembly is a rotating centrifuge in which the wastewater stream is evaporated at low pressure. A rotary lobe compressor then compresses the vapor and condenses it on the opposite side of the evaporator surface, conserving the latent energy. The mechanism is the same one a salad spinner uses to throw water onto the bowl wall, with the extra twist that the heat released when vapor turns back to liquid is recycled to evaporate more urine. For the underlying physics, see the article on phase changes and latent heat.

Lowering the pressure matters because liquid water boils at a lower temperature when the pressure above it drops, the same relationship that makes water boil at a cooler temperature on a mountaintop. The 2024 paper tracks the vacuum in the Distillation Assembly's condensing volume in the range of roughly 39 to 45 millimeters of mercury, a small fraction of sea-level air pressure. With the pressure that low, the still can evaporate urine without cooking it.

What comes out, and what stays behind

The distillation leaves two products. The distillate is mostly water. The brine is concentrated urine: the dissolved salts, urea, and other solids that did not evaporate, plus the pretreatment chemicals. The brine is stored in a bellows tank called the Advanced Recycle Filter Tank Assembly. When the brine is concentrated to its limit, the tank is emptied and refilled with fresh pretreated urine to start a new cycle.

The distillate is not drinkable yet. It still carries trace volatile compounds that evaporate along with the water, such as ammonia, alcohols, and organic acids. Those are the Water Processor Assembly's job.

The design load and how it was tested by reality

The UPA was designed to process a nominal 9 kilograms (19.8 pounds) per day of urine and flush water, the load expected from a six-person crew, and to recover 85 percent of the water in it. The 2024 NASA paper reports that this target ran into trouble in 2009, when urine quality led to calcium sulfate precipitating inside the hardware. Recovery had to be cut to 75 percent for urine from the U.S. segment and 70 percent for urine from the Russian segment to keep the still from clogging.

Engineers traced the problem to the pretreatment chemistry. Urine is pretreated with chromium trioxide, an oxidizer that controls microbial growth, and an inorganic acid that keeps urea from breaking down into ammonia. In the Russian segment the acid is sulfuric acid, and sulfate can combine with calcium to form scale. The U.S. segment switched to phosphoric acid in early 2016, and the UPA returned to at least 85 percent recovery on American urine. Russian-segment urine, which still uses the original formula, remains at about 70 percent.

The same paper reports that between February 2023 and February 2024 the UPA produced 4,867 liters of distillate, with recovery from 70 percent for Russian urine to 87 percent for U.S. urine. Planners are targeting 90 percent for U.S. pretreated urine, but they say more analysis of returned samples is needed first.

Year	Average crew size (max)	Process runs	Processing hours	Urine processed (lb)	Brine generated (lb)
2020	5 (6)	245	1,657	5,355	1,197
2021	7 (11)	326	2,544	8,296	1,627
2022	7 (11)	345	2,786	9,278	2,105
2023	7 (11)	359	2,964	10,486	2,249

Table: UPA operating metrics, 2020 through 2023. Source: Williamson, Wilson, and Luong, ICES-2024-317, Table 1, NASA Marshall Space Flight Center.

The table shows a clear pattern. More people means more urine, and the brine share stays between roughly 20 and 23 percent of the urine mass over the four years. That brine is exactly what the next piece of hardware was built to attack.

The Brine Processor Assembly: closing the last gap

For most of the system's history, brine left over from the UPA was treated as waste. It was transferred to Russian containers and then to a tank on the Progress cargo vehicle for disposal. It still contained a meaningful fraction of water, which meant the total recovery was stuck below the 98 percent that NASA set as its goal for long-duration missions.

The Brine Processor Assembly (BPA) was installed on the station in March 2021 as a technology demonstration. Instead of heating the brine, it holds it in a membrane-based bladder that lets water vapor pass through while blocking salts and solids, a process known as membrane distillation. Warm, dry air blown across the membrane carries the water vapor away as humid air. That humid air is then collected by the cabin's humidity control system, the same path used for crew breath and sweat, and rejoins the water loop as condensate.

In its 2024 paper, NASA reported that the BPA has achieved about 98 percent water recovery overall and that the BPA has completed 37 dewatering cycles since it was first activated in early 2021. In the reporting window from February 2023 forward, the BPA handled 18 dewatering cycles and processed approximately 396 liters of urine brine, of which an estimated 326 liters of water returned to the humidity condensate stream. The 2023 announcement attributes the jump from the earlier 94 percent to this hardware.

Membrane distillation sounds exotic, but the idea is familiar. A good rain jacket stops liquid water from coming in while letting sweat vapor out through its fabric. The BPA's bladder does the same job on a smaller scale: liquid brine stays inside, and only water vapor crosses the membrane, pulled along by the warm, dry air on the other side. The membrane, not gravity, is what keeps liquid and vapor apart, which is why the approach suits microgravity.

There is a subtle distinction worth keeping straight. The BPA by itself recovers water from the brine; it is the combined system that reaches about 98 percent. The 2024 paper separately states that the BPA raises recovery of water from urine brine beyond 95 percent, with early estimates of 97 to 98 percent recovered from urine overall. Different sources round and frame the figure differently, so the safest statement is that the total station-level recovery is "about 98 percent" in NASA's own wording, as demonstrated during the technology demonstration.

Inside the Water Processor Assembly: five barriers before the spigot

Distillate from the UPA and humidity condensate from the cabin both flow to the Water Processor Assembly. The WPA is a series of barriers, each tuned to a different kind of contaminant. The sequence, drawn from the 2024 and 2025 NASA papers, runs like this:

1. Gas removal and coarse filtering. Water first passes through a Mostly Liquid Separator that removes free gas, then through a 300-micron External Filter Assembly that captures biofilm and debris from the wastewater tank. A micron is one-millionth of a meter; 300 microns is about the thickness of three sheets of paper.
2. Particulate filter. A 0.5-micron filter protects the beds downstream from particles.
3. Multifiltration bed. This cartridge contains adsorbent (such as activated carbon, which holds contaminants on its surface) and ion exchange media (which swap unwanted dissolved ions for harmless ones). Before July 2019 the WPA used two beds in series. It now uses one.
4. Catalytic reactor. Volatile organics that the bed does not capture are oxidized at elevated temperature in the presence of oxygen. This step also kills microbes. Excess oxygen and gaseous by-products are then removed by a gas-liquid separator.
5. Ion exchange bed. The last scrubber removes soluble by-products from the reactor and adds iodine to the water as a biocide.

The catalytic reactor works on the same principle as the catalytic converter in a car, where a hot catalyst helps oxidize pollutants into harmless gases. If that comparison is new to you, the article on how catalysts speed up reactions explains why a catalyst can do this at temperatures where the reaction would otherwise barely move.

Why iodine, and why it is removed again

Iodine goes into the water so it stays free of microbial growth while it sits in tanks and pipes. The ES&E article explains that NASA uses iodine rather than chlorine because it is less corrosive and easier to transport to the station. However, iodine is not something NASA wants crew members consuming without limit. According to the 2024 paper, the point-of-use dispenser includes an activated carbon and ion exchange cartridge, abbreviated ACTEX, that filters out the iodine before water reaches the crew's cup. The iodine concentration in the stored water is tracked through samples returned to Earth, because it cannot be measured in real time. The paper reports readings around 2 milligrams per liter in late 2023 samples, well within the potable water specification.

Wearing out: how long the beds last

Components of the WPA have finite lives. The multifiltration bed installed in April 2022 processed 23,490 pounds of wastewater (about 10,650 kilograms) before it needed replacement in October 2023. The ion exchange bed is tracked by conductivity, a measure of how many dissolved ions the water carries. Think of the bed as a sponge with a fixed capacity: it soaks up ions until it is full, and once it is, ions "break through" and conductivity rises. Because the bed also supplies the iodine, it is sized so that the iodine does not run out before the ions do. The 2024 paper's history chart shows the ion exchange bed being swapped several times since 2008, and projects the currently installed bed to reach breakthrough at around 51,000 pounds processed, expected in June 2024.

Between February 2023 and February 2024, the WPA produced approximately 8,408 liters of product water. The UPA's lifetime total as of February 5, 2024 stood at 36,552 liters, according to the same document, produced with about 6,005 pounds of replacement hardware launched over the years. A single launch of 36,552 liters of water would have weighed over 36 tonnes before packaging.

What can go wrong: failures that taught the system

Recycling water in orbit is not a set-and-forget job. The 2024 paper is candid about anomalies, and each one teaches something about designing for deep space.

Problem	What happened	Lesson
Calcium sulfate scale (2009)	Distillation hardware clogged, forcing recovery down to 70 to 75 percent	Pretreatment chemistry is as important as hardware; U.S. side switched to phosphoric acid in 2016
Poorly pretreated urine (June 2023)	About 5 liters of badly stabilized urine from a Russian-filled container raised pressure in the still; a second batch of roughly 6.7 liters followed a month later	Inspect inputs visually; recover by flushing with extra pretreatment solution and replacing the brine filter
New pump technology (2023)	A scroll-type purge pump demonstration worked for about 500 hours, then lost its ability to hold vacuum	Test new hardware on the station before relying on it far away; engineers reverted to the proven pump
Sabatier reactor outage (since October 2017)	The carbon dioxide reduction hardware that yields water as a by-product stopped working	Replacement hardware was planned; the loop runs without it, but with less water returned
Wearing parts	The UPA's fluid pump passed 9,000 operating hours against a predicted life of 4,000	Increased crew size and duty cycle shortened margins

Table: Selected anomalies in the ISS water system, from NASA's 2024 status paper (ICES-2024-317).

The poorly pretreated urine episodes are instructive. When the first batch of about 5 liters arrived, the still's condenser vacuum pressure rose rapidly. Ground controllers commanded the unit into standby before it reached a fault limit. Visual inspection of the Russian-filled storage container showed degraded urine with apparent biomass growth. Engineers recovered by replacing the brine filter and transfer hose, then cycling an extra-strength mix of pretreatment fluid through the loop. The paper credits quick recovery, with little or no long-term damage to the hardware.

The pump demonstration shows a second kind of lesson. The scroll pump reached its target vacuum pressure in about 15 seconds, compared with roughly 1.2 minutes for the older peristaltic pump, a saving of about 80 percent in minimum runtime. Nevertheless, after about 500 hours the new pump lost its ability to hold vacuum, and the crew reverted to the baseline hardware. Faster is not better if it fails earlier, and that is exactly the kind of result a crew on a Mars transit could not afford to discover on its own.

Is the water safe to drink?

The usual assumption is that astronauts drink "yesterday's urine." That oversimplifies a multistep chain. Williamson, speaking to the ES&E article, said plainly that the crew is not drinking urine. The chemistry supports that: distillation separates water from salts and urea, the filtration beds adsorb remaining organics, the catalytic reactor oxidizes trace compounds, and sensors check the purity of the water before it is released. Water that fails the check is reprocessed.

Is it cleaner than tap water on Earth? That claim appears often, but this article does not make it, because a single meaningful comparison would need contaminant-by-contaminant data from both sources. What the NASA papers do establish is that the water must meet a potable water specification, and that iodine, organic carbon, and conductivity are tracked against limits. One example from the 2024 paper: when a siloxane-related compound called dimethylsilanediol loaded onto and then released from the multifiltration bed, total organic carbon rose in the product water. Upgraded sorbent and an added filter have since let the bed run to the end of its life. That is the sort of detail that suggests a system under active monitoring rather than one assumed to be clean.

The humor, and why it persists

Spaceflight culture has long joked about the loop, with the usual line being that yesterday's coffee becomes tomorrow's coffee. The joke persists because the engineering behind it is solid enough that people feel comfortable making light of it.

How water recovery ties into oxygen and the rest of the station

Water in orbit is also the raw material for breathable air. The Oxygen Generation Assembly takes potable water and splits it with electricity from the station's solar arrays into hydrogen and oxygen. Oxygen goes into the cabin. Hydrogen was historically fed to a Sabatier reactor, which combines it with carbon dioxide removed from the air to make water and methane. The water returns to the loop, and the methane is vented overboard.

This coupling explains why NASA engineers talk about the water loop as part of a "system of systems." A failure in one part, such as the Sabatier reactor outage described above, reduces how much water returns, and that change reaches the rest of the system. A separate piece on how electrolysis splits water explains the physics of the oxygen side.

Why this matters for missions beyond low Earth orbit

On the station, a cargo spacecraft can bring more water if something breaks. On a long deep-space mission, nothing can. Every percentage point of recovery changes how much make-up water a crew has to carry from the start. If a system recovers 90 percent, a crew of four using 3.8 liters per person per day needs to bring about 1.5 liters of fresh water per day for the whole trip. At 98 percent, the figure falls to about 0.3 liters per day. Those are simple calculations from the daily consumption figure above, not NASA mission-design numbers, but they show why the last few percentage points are worth years of engineering.

NASA treats the ISS as a testbed for exactly this reason. Williamson's 2024 paper describes the BPA's successful demonstration as a critical step before crewed missions beyond low Earth orbit, "because of the necessity to recover as much water as reasonably possible due to the launch costs for water and the absence of resupply capability." For the planning side of the Moon and Mars programs, see the articles on the Artemis program and Mars settlement engineering.

There are also gaps between what the station does and what a deep-space vehicle will need. The ISS system treats urine and humidity condensate. The 2023 ES&E coverage and the NASA papers describe a loop built around urine and humidity condensate, and a future long-duration habitat may have to handle more varied wastewater, such as washing water. The station also has a luxury a Mars ship will not have: regular crew rotations and cargo flights that let engineers swap failed components. NASA has begun testing parts designed for crew-level maintenance, such as the scroll pump assembly and a new Urine Transfer System, because repair at the component level is much cheaper than replacing a whole rack.

Upgrades aimed at the exploration era

Several hardware changes in the 2024 paper show where the technology is heading:

• The Urine Transfer System automatically routes urine from multiple toilets and storage containers into the processor, reducing crew time.
• A U.S. designed toilet (the Universal Waste Management System) arrived in December 2020 and has since seen several restarts and check-outs. It is intended for the Orion spacecraft.
• The eXploration Potable Water Dispenser, activated in September 2023, replaces a microbial filter with an ultraviolet reactor for point-of-use microbial control. It is the first use of ultraviolet light for that purpose in a station water dispenser.

Common misconceptions about the system

"Astronauts drink raw urine." No. The process removes salts, urea, and most organic compounds, then adds iodine and screens the output before anything reaches a drink pouch. The more accurate description is that the loop reclaims the water from several wastewater streams.

"The system runs itself." It does not. Filters and beds need replacing, pumps wear out, and in the case of the poor pretreatment episodes, the crew and ground teams had to intervene. The system is monitored by sensors, and ground controllers can command it into standby.

"98 percent means the station never needs water from Earth." Not quite. Some water leaves the loop with brine residue, food packaging, and experiment hardware, and contingency reserves such as the 818-liter minimum are replenished from the ground. The aim is to cut dependence on resupply, not to eliminate it.

"Closing the last two percent is trivial." The remaining loss is the hardest part. Brine solids still hold some water, the Russian-segment urine still recovers at only about 70 percent in the distiller, and pretreatment chemistry constrains what the hardware can handle. Each is an active engineering problem.

The takeaway

The station's water loop works because it treats wastewater not as one problem but as several, each answered by a different tool: a spinning low-pressure still for urine, membranes for brine, filter beds for dissolved contaminants, a hot catalyst for trace organics, and iodine for storage. The result, about 98 percent recovery in NASA's own wording, took more than a decade of fixes, from the 2009 scale problem to the 2016 chemistry change to the 2021 brine processor. For anyone planning a crewed journey away from Earth, the most useful lesson is less about the headline percentage and more about the failures documented along the way, because those are the ones that a crew with no cargo capsule would have to solve alone.

---

Read next:

• How Lagrange points hold space telescopes in place
• How spacecraft survive extreme temperature swings
• Why solar sails can move without fuel
• The ISS Explained: Living in Space

Sources

• Williamson, J., Wilson, J. P., and Luong, H., "Status of ISS Water Management and Recovery," ICES-2024-317, 53rd International Conference on Environmental Systems, July 2024, NASA Marshall Space Flight Center (NASA Technical Reports Server).
• Chen, L., Reuland, C., Rivas, J., Misek, W., and Rangel, A., "Unveiling the System-of-Systems Complexity in Regenerative ECLSS: Insights from ISS," ICES-2025-125, July 2025, NASA Johnson Space Center.
• "NASA's urine brine processor pushes water reclamation to 98% on International Space Station," Environmental Science & Engineering Magazine, July 4, 2023.

FAQ

Q: How much of the water on the ISS is actually recycled? A: NASA reported in 2023 that the station's water recovery, including the Brine Processor Assembly, reached about 98 percent, up from about 94 percent before the brine processor. The BPA has operated as a technology demonstration since March 2021.

Q: Is recycled space station water safe to drink? A: NASA says the crew is not drinking urine. The water is distilled, filtered through adsorbent and ion exchange beds, oxidized in a catalytic reactor, dosed with iodine, and checked by sensors. Water that fails the purity check is reprocessed. Iodine is filtered out again at the drinking dispenser.

Q: Why can't astronauts just bring enough water from Earth? A: Launch mass is expensive, and each person needs about one gallon (3.8 liters) a day. A crew of seven would need roughly 10,000 liters a year. On missions beyond low Earth orbit there is no resupply option at all, so recovering water is a requirement.

Q: What happens to the brine left over after distillation? A: Brine used to be disposed of. Since March 2021, the Brine Processor Assembly has held it in a membrane bladder and blown warm, dry air across it. The water vapor is captured by the cabin humidity system and returns to the loop, while the solid residue is discarded.

Q: Does the same technology apply to Mars missions? A: NASA describes the ISS system as a demonstration for exploration missions. It treats only urine and condensate, though, and a transit vehicle will need to handle more waste streams and be repairable by the crew, which is why NASA is testing component-level maintenance on the station.

Q: Why does the Urine Processor Assembly spin? A: Without gravity, liquid and vapor do not separate on their own. Spinning the Distillation Assembly pushes liquid into a thin film on the drum wall, where it evaporates at low pressure. The vapor is compressed and condensed on the outer side of the same surface, which recycles the latent heat.

Q: Why does NASA use iodine instead of chlorine? A: According to engineers quoted in Environmental Science & Engineering Magazine, iodine is less corrosive and easier to transport to the station. It is added as a biocide to keep stored water free of microbes and then removed by a filter cartridge at the dispenser.