Atmospheric Water Generation

Atmospheric water generation is not one technology.

Every system that pulls water from air has to do two things: capture moisture, then release it as clean liquid water. The sorbent chemistry that does the capturing is now well understood and shared across the leading approaches. What separates one platform from another is the engine that drives the release. That single choice decides where a system works, whether it runs around the clock, and how much it can produce.

7% RHLowest verified operating humidity

2 outputsPure water and inexpensive dry air

−20°C to 55°CValidated operating range

How AWG works

Two jobs, and only one of them is contested

Air everywhere carries water vapor, even over a desert. An atmospheric water generator does two things in sequence. First it captures that vapor, either by cooling the air until the water condenses, or by holding the vapor on a sorbent material. Then it has to regenerate: release the captured water as clean liquid so the cycle can repeat.

Job 1

Capture

Modern sorbents, including metal-organic frameworks, can pull water from very dry air. This chemistry is foundational science, it is increasingly mature, and the leading developers draw on the same body of work. Capture, on its own, is no longer the hard part.

Job 2

Regeneration

Releasing the water back out of the sorbent, continuously, in the field, at low cost, is the hard part. The method a system uses to do this is its regeneration engine, and it determines real-world energy use, reliability, climate range, and how far the system can scale.

The useful way to compare AWG systems is by their regeneration engine, not by their marketing. The four platforms below are grouped by how they release water, because that is the variable a buyer actually feels.

The landscape

Four platforms, defined by how they release water

Each approach is legitimate in the setting it was built for. The differences are architectural, not a matter of one company being smarter than another.

Condensation

Mechanical vapor compression

Who uses it: Watergen, Akvo, and more than 100 others

Cools incoming air below its dew point so water condenses on a cold surface, the same principle as a dehumidifier. It is mature, well understood, and effective in humid air. Most AWG products on the market today work this way, and it is by far the most crowded part of the field. Its limit is physical, not fixable by engineering: below roughly 40 to 60% relative humidity there is too little vapor in the air to justify the energy of chilling it, so output falls away and then stops. That rules out most of the arid and semi-arid regions where water stress is worst. It also produces water only.

Humid climates only ~40 to 60% RH floor Single output Efficiency ceiling

Sorption, released passively

Sorbent plus sun or ambient heat

Who uses it: SOURCE (formerly Zero Mass Water)

A hygroscopic sorbent captures vapor and solar heat releases it during the day, with little or no grid electricity. SOURCE has commercialized this as a rooftop hydropanel that raises the dew point inside the panel until the water condenses on its own, and it is a genuinely good fit for off-grid homes and remote sites. The limit is throughput. Tying the release to the daily solar cycle holds a panel to roughly one capture-and-release cycle per day and a few liters of output, and makes production depend on the weather. It serves a household, not a city or a factory, which is why developers chasing commercial and industrial volume have moved to powered, active regeneration.

~1 cycle per day A few liters per day Off-grid / residential Single output

Sorption, released by a powered engine

Active regeneration, varied engines

Who uses it: AirJoule, Aquaporo, Uravu, Atoco

This is where most serious AWG developers now sit, and it is the right move: a powered engine releases the water on demand, so a system can run around the clock and reach into drier air than condensation. The engines differ, and so do their tradeoffs. AirJoule pairs a metal-organic framework with a vacuum-swing engine; it works, but the sealed chambers and pumps are maintenance-heavy and its best efficiency leans on co-located industrial waste heat. Aquaporo's ALMA system applies direct heat plus a separate chiller; it reaches low humidity, near 15% RH, but spends a lot of power to do it. Uravu takes a different route again, regenerating a liquid desiccant with low-grade or waste heat; it is efficient in humid air, but the regeneration energy climbs steeply as humidity falls, which is why this approach suits warm, humid regions. Atoco, built on the same foundational MOF science that the rest of the field draws on, has moved from its original passive design to an active system, but has not disclosed how it regenerates, which leaves its performance claims unverified for now. The thread running through this group is energy: most of these designs spend energy to release the water and then let it escape, so they draw heavy power or depend on an outside heat source. The output is real; efficiency and independence from a particular site are the open questions.

Continuous operation Reaches ~15 to 30% RH Energy-heavy or waste-heat-tied Single output

The WaHa Vaporator®

Sorption with a closed-loop heat-pump engine

Active regeneration, sealed and self-powered

WaHa uses active regeneration as well; the difference is what happens to the energy. A sorbent captures vapor, and a sealed, closed-loop heat pump drives the release, recovering the latent heat of condensation and reusing it to power desorption. Regeneration energy stays roughly constant across a wide range of conditions instead of collapsing when the air gets dry, and the system needs no vacuum and no outside heat source. It runs on grid power, on renewables with storage, or on waste heat where available, with no dependence on any one of them. Because the engine, not the weather, controls the cycle, the WaHa Vaporator® runs continuously, operates down to 7% RH (roughly 1 g/kg), and holds output across −20°C to 55°C. The same thermodynamic engine yields two products from one energy input: pure water and inexpensive dry air. The architecture is sorbent-agnostic, so the system is not tied to any single material supplier.

Down to 7% RH Runs 24/7, weather-independent Water + dry air Grid, solar, or waste heat No vacuum seals Sorbent-agnostic

Side by side

The platforms compared

Grouped by architecture, with the companies working in each, because every product inherits the physics of its platform.

What matters to a buyer	CondensationWatergen, Akvo, +100	Passive sorptionSOURCE	Active sorptionAirJoule, Aquaporo, Uravu, Atoco	WaHa Vaporator®Heat-pump engine
Works in dry air	No, stops near 40 to 60% RH	Yes, but low throughput	Yes, typically to ~15 to 30% RH	Yes, down to 7% RH
Runs continuously	Yes, in humid air	No, tied to the daily cycle	Yes	Yes, 24/7, weather-independent
Energy story	Rises sharply as air dries	Sunlight or ambient heat	Heavy power draw, or depends on outside waste heat	Recovers its own heat; 0.24 to 0.45 kWh/L on grid, solar, or waste heat
Outputs	Water only	Water only	Water only	Pure water and inexpensive dry air
Field serviceability	Standard refrigeration service	Simple, but limited capability	Vacuum and specialized engines need trained service	Standard HVAC service, no vacuum, no exotic parts
Scales to industrial volume	In humid markets only	Not at continuous volume	Yes, but at high energy cost	Yes, modular across the product line
Best-fit use	Humid, temperate regions	Off-grid, humanitarian, small scale	Sites with cheap power or spare waste heat	Anywhere, including the driest water-stressed regions

Field-validated, not just demonstrated

The WaHa Vaporator® has been proven through multi-month pilots in UAE and Saudi desert conditions and in Swedish cold, run by enterprise and institutional partners, not in a one-off demonstration.

ExxonMobil/ TAQA/ EBD Paragon/ UC Davis/ Arizona State University

Why the engine is the difference

The moat is the engine, not the material

It is tempting to think the magic in next-generation AWG is the sorbent. The sorbent matters, and the chemistry behind today's best materials is genuinely foundational work. But that is exactly why it is not the differentiator: nearly every advanced AWG developer now builds on the same family of metal-organic frameworks, drawing on the same body of research. When everyone has access to remarkable materials, the material stops being the edge. And a remarkable material still does nothing useful until something releases the water it holds, reliably, in the field, at a cost that pencils out. That is the regeneration engine, and it is where the WaHa Vaporator® is built to win. A sealed, closed-loop heat pump that recovers and reuses its own condensation energy is what lets one architecture run in a Gulf summer, a cold northern winter, and a humid warehouse without changing its energy story. Because the engine is the differentiator, the platform does not depend on any single sorbent, and its advantage holds even as sorbent chemistry keeps improving for everyone. That sealed loop is the keystone of WaHa's granted patent portfolio, one of the deepest in atmospheric water; current counts and country coverage are maintained on our patents page.

The competitive question in AWG is no longer "can a material capture water from dry air." It is "what releases that water, continuously, anywhere, at scale." That is an engineering question about the engine, and it is the one WaHa set out to answer.

Proof, not claims

See it on a map, city by city

Claims are easy; maps are harder. We built an interactive tool that tests where each platform meets a utility-grade reliability standard: at least 90% of rated daily production on at least 90% of days, computed from four years of public climate data. Adjust the standard's strictness yourself, search any city on Earth, and see the technologies judged by one identical test. We publish where our machine works and where it does not, and we have proposed this method as an open standard for the industry.

Explore the viability map →

For evaluators

What to ask any vendor

Six answers define a water machine. We have proposed an open industry standard that makes every vendor give them the same way; until it is adopted, ask directly. These questions apply to every architecture, and the answers separate a water supply from a water gadget.

Where does it work?

At what humidity does output begin to fall, and where does it stop entirely?

Ask in grams of water per kilogram of air, not relative humidity: relative humidity swings with temperature even when the water in the air has not changed. Ask for field data, not a lab best case.

What share of days at your site would the machine deliver at least 90% of its rated production, across all four seasons of a real year?

Ask to see it computed from climate data, not asserted.

What does the machine need from its site besides air and electricity?

Waste heat at a particular grade, a cool heat sink, direct sun? And what happens to output when the site does not cooperate: a cloudy week, a dry season, a 45 °C afternoon?

What does it truly cost to run?

What is the energy per liter counting every input that crosses the machine boundary?

Electricity, thermal input and its source, heat rejection and its sink. If any input is described as free, ask what the figure would be if you paid for it.

Beyond the headline number, what auxiliary loads run continuously?

Fans, pumps, vacuum systems, controls, and how they trend across a year as components wear and filters load.

What comes out of the tap?

What is the water quality at the point of dispense, after storage, not at the condenser?

Mineral content, disinfection method and residual, and microbial safety through storage and use, including Legionella.

Does the machine show up?

What is the longest continuous field deployment, independently verifiable, and what fraction of rated output did it deliver?

How did availability and output trend over that run? A machine that passes the climate test on paper but cannot run continuously in the field is still a supplement.

What does ownership look like?

What is the capital cost per daily liter of rated output, and what does a year of upkeep look like?

Every consumable, its replacement interval, and annual service hours, so you can price your own labor. And does the machine produce anything else of value, such as dry air or cooling, and is that counted in the economics?

How does it live where you put it?

What is the noise level in dB(A) at rated production, at a stated distance?

Not standby. Machines have been shut down by their neighbors, outdoors, for failing this one. Ask also about heat rejected into your space, refrigerant type and global warming potential, and end-of-life.

The bigger picture

One standard family: what is live and what is coming

Where a machine works is the first of four questions every buyer should get answered the same way from every vendor. We have proposed an open standard family for the industry, and we hold ourselves to it first.

Part I, where does it work? Live

The interactive map, the full methodology, and the reference results are published and reproducible. Where It Works.

Part II, what does it truly cost to run? Defined

Every energy flow crossing the machine boundary counts, thermal included; free heat changes the invoice, not the thermodynamics. The declaration format is published in the draft standard.

Part III, what comes out of the tap? In drafting

With independent laboratories. Tested at the point of dispense, not the point of production.

Part IV, does the machine show up? Defined, first field record published

17 months and 10,143 hours of continuous operation in the UAE at 95% of nameplate production.

Two disclosure schemas: cost of ownership; environment and siting. Formats defined

So vendors report and buyers compare on identical terms.

See the platform that works where the others cannot

The WaHa Vaporator® Oasis-24 series is open for pre-order in outdoor and integrated configurations, with delivery in 2027. For larger systems and pilot deployments, talk to our team.

Pre-order the Oasis-24 Talk to sales about a pilot