Atmospheric Condensing:
A Proposed Solution to Global Thirst

(Submitted to the San Diego Foundation for consideration
in the 1999 Blasker Award for Environmental Science and Engineering

Patrick G. Salsbury

Founder, Reality Sculptors

First Draft: Oct. 21, 1998


Approximately one billion people on this planet do not have safe supplies of drinking water. This, coupled with poor sanitation, contributes to approximately eighty percent of the world's sickness1. Forty percent of the world's population face water shortages2, despite the fact that water covers approximately three-fourths of the planetary surface. It flows around us, through us, and falls from the sky. It is perfectly abundant, yet not always in the form we want, nor in exactly the place we want. Sometimes there's too much, oftentimes too little.

This research attempts to outline methods for extracting abundant supplies of fresh, drinkable water from the atmosphere. The ultimate goal is to provide a reliable, renewable, and ecologically friendly source of drinking water at just about any habitable location on the planet. This can be achieved without large centralized systems, without economically costly public-works projects, and without the ecologically costly burning of fossil fuels, if so desired.

Most people are familiar with the sight of a glass of ice water. It is cold, cool, refreshing, and covered with condensation on the outside. That condensation is forming due to moisture in the air coming into contact with the cold surface of the glass. As the air cools, the gaseous water molecules slow down, begin to accumulate, liquefy, and drip down the glass. This basic principle of water will be explored in the following text, and ways will be discussed to maximize output of useful water, and turn it to the needs of humankind.

First, it's necessary to dispel a prevailing myth about water: There is no shortage of water on this planet. There is, however, a shortage of drinkable water. The oceans and seas provide the largest repository of water, approximately 97% of the total water on earth.3 This water is, of course, salt water, and undrinkable by humans, so we must intercept our water from other parts of the hydrological cycle.

Unfortunately, most sources of fresh water are rather geographically confined. You can't, for example, easily get water from the Great Lakes if you happen to be in the Sahara Desert. Thus, most regions must rely on their own local geography to catch rainfall, and divert it into streams, rivers and lakes. This water will also be retained in the local soil by plant rootmass and underground aquifers, eventually making its way back to the sea. (See Figure 1)

JPEG image of the Hydrological Cycle

Figure 1- The hydrological cycle, showing the Earth's water inventory, annual movements of water through the cycle (black numbers), and amounts of water stored in different parts of the cycle (blue numbers). Note the overwhelming dominance of ocean water in the inventory. All quantities shown are x 10^15 kg. Note also that 10^15 kg water = 10^3 km^3.4

Studying the data in Figure 1, it becomes clear just how much water there actually is in the atmosphere. Each unit in the above graphic represents 1000 cubic kilometers of water. That means that there are approximately 13,000 cubic kilometers of water "resident" in the atmosphere at any given point in time. There's always some evaporating someplace, and falling as rain or snow someplace else, but on average, there is an immense amount just floating around.

Looking next at Table 1, we see a breakdown of the water supply on land. These numbers are in the same units of measure as above. To put the atmospheric water supply in perspective, there is thirteen times as much fresh water in the air as there is in all of the rivers and streams of earth. The atmospheric water also totals about one-tenth the total of all freshwater lakes on the planet. And due to the constant shifting of the hydrological cycle, any water that we may draw out of the air in one place will be replaced by evaporation someplace else.

Table 1 Water on land (x 10^15 kg).5

Rivers and streams


Freshwater lakes


Salt lakes and inland seas


Total surface water


Glaciers and ice-caps


Soil moisture and seepage




Total on land



The question, of course, is "How do we get at all that water?" And the answer is, "There are multitudes of ways." Rain is simply water condensing around minute particles in the air. Plants at night catch fog and dew, which drips down to their roots. A mature redwood tree can draw up to 500 gallons of water per night from the foggy coastal air of California.6

Certainly, if Nature can do this, so can humans. Going back to our glass of ice water, we see the basic principle: Have a cold surface, and a flow of humid air going over it. A glass doesn't have much surface area exposed, yet it still produces a fair amount of condensation.

So, to reduce it to basic principles, we want something that's very close to zero degrees Centigrade (the freezing point of water), while allowing the water to remain liquid. (Otherwise, we're left with a defrosting problem.) We want to maximize surface area, just like the giant redwoods do, and to maintain good airflow over and around that surface, so as to keep a flow of fresh molecules that may catch and condense on the apparatus. (See Figure 2)

JPEG image of the schematics of an Atmospheric
Condensing System

Figure 2 Basic Flow of an Atmospheric Condensing System

It is also important to keep energy input to a minimum. It's certainly not difficult to make something cold (we've had refrigeration technology for about a century and a half), but it's desirable to do it as efficiently as possible. Energy-intensive systems tend to pollute, are more expensive, and will be less likely to be deployed in poor Third World regions where they're needed most.

Ideally, a device can be developed that is small, inexpensive, portable, solar powered (solar-electric and solar heat can both be utilized) and able to be left unattended for weeks, months, or years.

The author has built a countertop prototype of such a device. This was a proof-of-concept model, and was by no means self-sufficient or solar powered. However, it was able to draw water out of the air in varying amounts, depending on ambient temperature, humidity, etc. It was built from a discarded heater core (a small radiator with lots of surface area) from an automobile, through which ice water (mean temperature of ~1 degree Centigrade) was cycled with a recirculating pump from an adjacent reservoir.

For the purposes of this test-run, ice water was chosen as a convenient method of keeping the temperature in the 0-1 degree Centigrade range. It avoided the problems of frost and freezing, while keeping the metal surface of the radiator well below the dew point. It could certainly be done with conventional refrigerant mixes, but would need to be optimized to prevent freezing.

Tests were run on a variety of days during the summer and early fall of 1998, in the Santa Cruz Mountains of California. At temperatures ranging from about 15 to 32 degrees Centigrade (~60 to 90 degrees Fahrenheit), and relative humidity of between 40-70%, this prototype was able to condense water at rates from 1 cup to 1 quart of water per hour. The heater core had dimensions of approximately 6"x8"x2", but due to the hundreds of folds in the copper metal, had a surface area calculated at just over 1000 square inches, or about 7 square feet.

Initial tests revealed that, while water droplets would readily condense on the cold metal, little to none would drip off into a catch basin. Putting a small DC fan near the apparatus to direct a flow of air across the coils produced copious amounts of water, as noted above. Further research and experimentation will be needed to achieve an optimal balance of airflow, temperature control, insulation, etc.

With the results shown by this prototype, it is possible to calculate what scaling could do for the total amount of water collected. Taking only 50% of the observed results, to factor in variations that will arise based on temperature, relative humidity, and the like, we still see a condensation rate of approximately 1-pint per hour or 3 gallons of water per day from the small countertop model. Scaling this by a factor of 10, we get a device smaller than an average microwave oven with 10,000 square inches of surface area, and a fresh water output of approximately 30 gallons per day.


Of course, as with most things, "there's more than one way to do it". The countertop prototype used electricity from the wall socket to run the fan, and ice from the household freezer to keep the reservoir cold. This was done mainly because those resources were readily available. This won't hold true in most places, though, so further refinement is necessary.

Ironically, it generally takes heat to make cold. Electric refrigerators use mechanical compressors to compress refrigerant gasses. As those gasses expand, they cool off. The waste heat from the compressor is vented to the surrounding area. Propane refrigerators work slightly differently, doing away with the compressor, and using the heat of burning fuel to induce a phase-change in a refrigerant. Solar powered refrigerators can work on either principle. Solar-electric may use electricity to run a "standard" (compressor-driven) refrigerator or solar-heated pipes and refrigerant materials can be employed in much the same way as a propane refrigerator.

Using solar power to generate refrigeration seems like a sensible approach, since there's so much incident solar energy hitting the planet daily. A square meter of the earth's surface receives (on average) up to one kilowatt of power per day.7 Most of this ends up as heat. There are little to none of the attendant problems of pollution that arise from using petrochemicals, and the areas with the most sunshine are often the hottest and driest, thus likely to have a need for a steady supply of fresh water.

The Real Goods Corporation sells a solar-powered icemaker that uses no electricity, gas, or Freon. It uses the Ammonia Absorption Cycle, which has been in use since the 1850's. This ice-making machine uses solar heat to store energy in the form of high-pressure ammonia. At night, water is added to the machine. As the ammonia evaporates back into the main chamber, it provides enough cooling power to make ice. Depending on the model of the machine, it can produce between 10 and 1000 pounds of ice per night.8 Referring back to Figure 2, this solar powered "cold machine" would be able to keep the water in the insulated tank hovering around 0 degrees C, and this chilled water would be circulated through the radiator manifolds to create cool surface area for water to condense on.

Cool, drier air is a by-product of the condensing process. In fact, air-conditioners at work in our homes, offices, and automobiles are running condenser setups, but they don't bother to collect the water. By integrating both of these goals, we can arrive at a device that not only conditions the air, but also provides supplies of fresh drinking water.

Large universities and some municipalities use the same schematic as shown in Figure 2, but on a macro-scale. They will have a "chilled water plant" that pipes cold water to the various buildings on a campus, and this water then passes through manifolds to cool the air, and remove moisture. Utilizing solar heat and the Ammonia Absorption Cycle, we can get the best of both worlds, and avoid eating up lots of power generated by fossil fuels.

Another surprising example of the results possible with condensing is seen in the work of Professor Carlos Espinoza of the Northern University of Chile. He created a prototype condenser unit consisting of 1300 vertical artificial fiber filaments which provided surface area for water droplets to form on and drip into a holding tank. The area in Chile where he was testing the device is very interesting in that it has a heavy coastal fog coming in from the Pacific Ocean, very similar to the fog in the San Francisco Bay Area which the redwoods use for their condensation. Even more interesting is that this region is one of the driest desert regions on planet earth. The average annual rainfall in this area is less than 3 millimeters.9

Professor Espinoza's condenser ran for 2 years, yielding an average of 4 liters (7 pints) of water per day. The device had a condensation surface area of about 1.3 square meters and a .3 square meter "footprint" on the ground. It condensed this 4 liters of water per day without using any power whatsoever. Obviously, this design could be scaled to increase the amount of surface area (and thus water collected), but his results as they stand are quite startling. 4 liters per day in a .3 square meter area comes out to 110 centimeters of water per year. Based on area, this rate is about 41% more than the average world rainfall of 78cm/year.10

The "Oasis Machine":

By tying together the various power and cooling technologies mentioned above, we could build a machine that just sits in the sun, drawing the heat and power it needs to create the cold necessary to run a condenser.

This condenser would be able to draw water from the air at night (when the dew point is lower, and air has a higher moisture density), and possibly during the day, depending on humidity level. Water could be dripping off of it 'round the clock, and collecting in a catch basin below. By using a pond-liner,11 that catch basin can be very large indeed, and water could be retained for future use, without seeping into the ground and disappearing. In essence, it would be like the fountains that are the centerpieces of plazas and villages around the world, but rather than water being pumped in or simply held there, it would actually be condensing there to be used by the people of the village.

Imagine, for a moment, some random spot out in the wilderness. By dropping a large pond-liner into a natural depression, and setting up one or more of these "Oasis Machines" adjacent to it, or on islands in the middle, water would begin to collect, and form a quasi-natural watering hole. Over time, (a few years) this oasis would grow to provide a communal gathering place for people, as well as a resource and habitat for animals, birds, and fish. It would function as its natural counterparts, but with the added benefit of being able to be placed, rather than having to rely on just the right natural circumstances. It could run unattended, drawing water from the air with power from the sun, and provide the steady supplies of water necessary for long-term forest communities to develop. If done with an eye towards aesthetics, the Oasis Machine could be disguised to look like a natural feature. Perhaps it might even appear as a large moss-covered rock that drips water into the pond...constantly.

By gaining the ability to create oases where desired, humans will be able to begin moving away from the traditional (and damaging) methods of securing a steady water supply. There will be a decreasing need to dam or divert streams and rivers, to flood valleys to make reservoirs, or to pump water up from underground aquifers at a rate faster than they naturally replenish. (Mexico City is a notable problem area for this, and the land in parts of the city is beginning to subside as the underground water is removed and the aquifer compacts. Damage to surface structures is also apparent. People of Egypt and the Sahara are pumping water out of non-replenishing aquifers.12)

The Gory Details:

As noted, there is still much work to be done. The end goal is to have at least two different classes of condenser. One should be large enough to handle the centralized needs of a Third World village or small rural community, and should have the ability to run from solar power, or some sort of fuel, in case electricity is not available. The other should be small enough for an individual home or family to use. It should priced comparable to other mid-range kitchen appliances, such as microwave ovens or bread machines, and output at least enough water to meet the drinking/cooking needs of a family. A third scale would be on the order of the "chilled water plant" designs already in use in some places, although this would entail a much larger investment in infrastructure.

There are many devices that could be used either "off-the-shelf" or with only slight modifications to set up a workable arrangement. The aforementioned ice-making machine from Real Goods has been deployed in African villages where there is no electricity. That machine has coils used to generate extreme amounts of cold in order to make ice. Modifying the coils to increase their surface area, and the ammonia mix in that system (so that it doesn't get quite so cold), should allow the reworked machine to draw of a steady supply of clean water from the atmosphere. Conversely, as noted above and in Figure 2, the machine could be used "as-is" to simply provide ice-cold water that was circulated through the condenser unit.

There are plenty of photovoltaic cells available on the market today, as well as windmills for generating power. There are super-insulated refrigerators and freezers. Most of the hard work is already done. What's needed now is integration of these various tools, and a way to deliver them to the people that need them most.

Since the people who need fresh water the most are also generally the poorest of the poor, relations with the various local governments in drought-stricken and underdeveloped areas will have to be established. By providing an "appliance", rather than proposing a massive, centralized system, we can sidestep some of the pitfalls inherent in traditional public works projects. Decentralized systems tend to be less prone to overall failure. Should one condenser fail in one house, just that house is out of commission. Should a water main burst in a city, thousands of families may be affected.

Using off-the-shelf technology is also desirable to achieve economies of scale. Once the initial research and development is completed, condenser units will need to be produced in quantities numbering in the tens or hundreds of millions, at least. (See the opening line of this article's Abstract.) To this end, having a design that is simple, inexpensive, easily replicable, and durable will be of paramount importance.

It is important to stress the durability aspect. All too often, manufacturers start with a robust product, and over the years begin to "weaken" the design, so that it needs replacement intermittently. This should not be the case with Oasis Machines. As noted earlier, there are currently approximately 1 billion people in desperate need of this technology, and about 2 out of 5 people on Earth at least suffer water shortages. There is no need to go down the road of "planned obsolescence" in this market. Aside from the ethical questions it would raise, there is simply no economic need. There is no shortage of people who can make use of this technology, and the Earth's population is not shrinking.

The production of a condenser and reliable supplies of fresh water is a standalone goal that addresses immediate needs, and can help to improve standards of living. It is also one step towards the larger goal of providing comprehensive solutions to the vast numbers of humans who don't have enough to eat, clean and safe water supplies, a roof over their heads, adequate power, enough education, or the other components that make for a healthy life on Earth.

The initial tests seem to bear out that the theory is sound. Every cold glass of ice water is a testament to the possibility of drawing water from the air. Now what's needed is more rigorous testing to try and maximize water output, minimize production costs, and package it in a manner that makes it accessible to people with no special training (or even interest) in atmospheric condensing. It needs to be as simple as a toaster: Plug it in, hit the button, and it runs.

Water, Water, Everywhere:

Once the problem of fresh water supply has been taken care of, there are still many other things to deal with, such as keeping the water pure, free of pathogens, etc. There are various filtration and purification systems that can be incorporated into the design. Most likely these would be in the storage or dispensing stages. Chlorinated water is a familiar solution, but other systems, such as ozone treatment or filtration through a household system are also possible.13

New possibilities open up, too. Once a steady supply of water is available, it becomes much easier to raise one's own supply of food. By incorporating hydroponics, bountiful crops can be raised with a minimal expenditure of water as compared to soil-based agriculture, with its heavy water losses due to evaporation, and topsoil losses due to erosion.14 Even in existing agricultural situations, mid-range condenser units can be placed on farm properties to collect and hold water in local ponds or storage tanks, and fed to the local fields without having to run miles of pipes to connect to a municipal supply. This opens up vast new areas to potential farming, areas that are far from man-made water supplies, and which don't naturally collect enough water to sustain farming. The area that Professor Espinoza was researching in Chile is about 160 miles wide, and 2000 miles long, running most of the length of South America. It is one of the driest desert regions in the world, yet by combining condensation technologies with hydroponics technologies, that area has the potential to become one of the most productive and abundant agricultural areas of the planet.15

As one last point of comparison, the average person in the UK uses 135 liters of water per day, as opposed to 10 liters per day in the Third World.16 Clearly, what is needed is not simply more water, but also a rethinking of how we use it, and devices that help to conserve what we do use.

Further Reading/Call for Participants:

The Reality Sculptors Project hopes to draw together talented and visionary people from around the world to address pervasive social problems such as food, water, and housing shortages. It is the intent of the Project to develop systems such as atmospheric condenser units that can be deployed around the world in individual homes, or in villages and towns, to supply people with renewable supplies of fresh water for drinking and agricultural needs.

Skills especially useful towards this end are:

Interested readers are encouraged to join the "clean-water" mailing list or any of the other Design Science lists in the Reality Sculptors Project that appeal to them. Complete instructions are available online at

References and Notes:

1 UN Development Program, as quoted in "Naked Body", Summer 1998. Printed by The Body Shop
2 Friends of the Earth, ibid.
3 The Open University - --"Seawater: Its Composition, Properties and Behaviour" 2nd Edition, 1995 , p. 7
4 Graphic and caption text - ibid., p. 7 (Originally called Figure 1.3)
5 Table reconstructed from Table 1.3 - ibid., p. 7
6 The author currently resides in the redwoods of Northern California. This factoid arose from conversation with a local historian and environmentalist.
7 Medard Gabel with the World Game Laboratory - "Energy, Earth and Everyone", 1975,1980, p. 137 -
8 John Schaeffer & The Real Goods Staff- "Solar Living Source Book", 8th Edition, 1994 Real Goods Corporation, p. 259 -
9 Dr. Allen Cooper - "The ABC of NFT (Nutrient Film Technique)", 1996 Casper Publication Pty Ltd., p. 88 -
10 ibid., p. 88
11 The Real Goods Corporation sells pond liners like this. They range in size from very small (bathtub-sized) to hundreds of square yards in area.
12 Lester R. Brown and the Worldwatch Institute - "State of the World - 1996" 1996 by Worldwatch Institute, p. 42
13 There are household filter systems available that are inexpensive and simple to maintain. They can sit in your refrigerator, on a countertop, or attach directly to your plumbing.
14 Patrick G. Salsbury - "Hydroponics and Housing for the 21st Century", 1996 -
15 Dr. Allen Cooper - "The ABC of NFT (Nutrient Film Technique)", 1996 Casper Publication Pty Ltd., p. 87 -
16 Water Services Association, as quoted in "Naked Body", Summer 1998. Printed by The Body Shop

Patrick Salsbury

Last modified: Friday, 06-Aug-2004 04:20:47 PDT