Evaluating the water-collecting properties of various materials in a low-cost dew condenser for plant growth in a simulated arid climate

Aurel Lazar

Spring Valley High School

120 Sparkleberry Lane

Columbia, SC 29229




Desertification has for a long time been plaguing our world. In recent years, exponential increases in human population have only exacerbated the situation, calling for new techniques and strategies to be employed to halt the spread of arid desert areas. One such approach involves the collection of water for vegetative growth. A plausible solution involves a wire farm in which water condensation forms on nocturnally cooled wires in fairly hot desert air, successively dripping from wire to wire until it reaches an underlying plant row. Various condensing substrates were analyzed for their water collecting abilities. When wiring was determined to not function, steel, rubber, plastic, and glass tubing filled with water were used as condensation substrates. It was hypothesized that water yield would increase with an increase in specific heat of the substrate, and subsequently that non-metals would produce a larger yield than metals. All the data collection was done in an environmental chamber set to averaged desert conditions, and the results for each section were analyzed using an ANOVA test, a Tukey test and a Correlation test. The ANOVA test revealed a significant difference between the water yields of the groups. The post-hoc Tukey test revealed a difference between glass and all groups, as well as rubber and all groups. A further correlation test showed that specific heat was not directly correlated with water yield. Wire farms may be a method of extracting water where it is not found, but future studies are needed to fully explore the concept.


One of the greatest environmental issues currently plaguing the natural world is the sporadic and random expansion of arid desert areas, a process known as desertification. This process that occurs naturally over long periods of time, has only recently been identified as a worldwide phenomenon. In the past, people tended to move to new land whenever they had depleted a certain area of its resources, and thus expansion of deserts was not an observable process. However, with the recent exponential rise in human population, heavy demands for food and a disappearance of habitable land have made it clear that human mismanagement of land can lead to complete desertification of an area (Dregne, 1986).

There are many factors that influence when and why desertification occurs. It is a common misconception that droughts cause desertification; rather, abusing the land in such cases when it is already stressed contributes heavily to land degradation when it would otherwise recover (Watson, 1997). Usually desertification is the result of wind erosion on the soil when all vegetation cover is removed. The most common reason vegetation is removed is by the grazing of animals, and when animals are fenced into small lots, overgrazing occurs. The animals consume the vegetation cover and trample the exposed soil with their hooves, heavily compressing it and making it impermeable to further plant growth. This contributes to making overgrazing the leading cause of desertification (Collins, 2001).

Sand dunes are commonly associated with deserts and are easily blown away to other areas. Without a vegetative stratum to stop it, airborne sand may accumulate in a previously loamy area. In other cases, sand dunes migrate and encroach upon villages and cities (Bell et. al., 2002). Modern physics cannot predict the movement of granular particles such as sand because each grain is unique and interacts distinctively with those particles around it when flowing. Most of the time the particles undergo saltation, or jumping to a different dune (Bell et. al., 2002). Such situations are aggravated by lack of a plant root system to keep the soil structure firm rather than unstable.

Vegetative cover becomes evident as an extremely important facet in preventing the further spread of desert systems. Many are simply unaware of the importance that plants contribute in stopping the advancement of deserts. However, in order to survive, plants depend heavily on water. In fact, arid climates such as deserts are known for having extremely little precipitation and for being dry.

Most people are unaware of the vast volume of water available in airborne atmospheric rivers (Nelson, 2003). A major problem is that though the water is there, it cannot be harnessed without a mechanism for reaching it. Usually, the water is high in the atmosphere, and any collectors must be built on mountains. In many places without water, people have created retrieval systems such as airwells, fog fences, desalinization plants, and other mechanisms for collecting large quantities of fresh water. The Zibold airwell consists of a large pile of stones that are supposed to generate water from diurnal condensation to be transferred into pipes (Kogan et al., 2003). Fog fences are built in areas with high quantities of fog, converting the fog to water to be used in everyday life (Nelson, 2003). In some areas where only salt water is available, desalinization plants are used to obtain fresh water, but the plants require a high input of fossil fuels that contribute to atmospheric pollution (Alekseev et al., 1998). The problem with systems such as these is that they are extremely large, taking up possible farmland, and are usually too expensive for the average desert subsistence farmer.

In order to properly combat desertification, new systems for water collection must be developed. Such a system must remain relatively inexpensive and be made with commonly found materials. One solution is based upon the concept of a wire farm, consisting of wires suspended above plant rows. Because of the vast temperature differences between day and night, the wires cool down during the night, especially if cloud cover is removed to expose the heat sink of the exosphere (Walters, 1999). In such a way, the substrate must be exposed to radiative cooling (Muselli et al., 2002). As daytime comes, the air surrounding the cooled wires becomes warm enough to allow the water vapor in the air to condense on the wires as dew if the wires are below the dew point (Walters, 1999). The droplets of water, if large enough, will drip onto the plant rows underneath them. To enhance the process, multiple wires could be used stacked on top of one another, allowing droplet size to increase as droplets fall from one wire to the next.

Determination regarding whether or not certain materials have a better capability at condensing water in the aforementioned manner is imperative. It is clear that some materials naturally have a higher tendency to condense water, and this may be due to the adherence of the water to the surface, the specific heat, or even surface area. An initial test used a variety of materials as the substrates for dew condensation. Utilized as wires, the materials were the independent variable, and included glass, aluminum, plastic, copper, and steel wires. The reason for the selection of the various materials revolved around the value of the material. Obviously, inclusion of commonly used materials is much better for practicality, than is obtainment of the best and most expensive materials. Because the volume remained the same for all materials, it was hypothesized that the material with the greatest mass and the material with the lowest specific heat would produce the most water. As the material becomes denser, and can remain colder for a longer period of time as the outside temperature increases, more condensation will be formed. As a result, it was expected that non-metallic objects would be much more functional.

After the first experiment, a second experiment was conducted utilizing tubing filled with a coolant rather than wiring. The coolant was used to raise the specific heat of the system to such a point that the system would remain colder for a longer transitionary time. Various materials, namely rubber, plastic, steel, and glass tubes, were compared in terms of water yield. It was hypothesized that the materials with the highest specific heats would produce the greatest water yields, and consequently that non-metals would produce more water than metals, as it is known that metals tend to have low specific heats.




Aluminum Wire

Steel Wire

Plastic Wire

Glass Wire

Copper Wire

Environmental Chamber


Test Tubes

Cork Stopper


Glass Tubing

Plastic Tubing

Rubber Tubing

Steel Tubing




IV: Wire Material
Copper / Aluminum / Plastic / Steel / Glass
6 Trials / 6 Trials / 6 Trials / 6 Trials / 6 Trials

DV: Water Yield (mL)

C: Humidity
Regulated Temperature Bounds
Testing location

Wire volume

IV: Tubing Material
Glass / Rubber / Plastic / Steel
6 Trials / 6 Trials / 6 Trials / 6 Trials

DV: Water Yield (mL)

C: Humidity
Regulated Temperature Bounds
Testing location

Wire volume


After obtaining the various substrates, their volumes were determined using the formula , allowing for a calculation of density. The environmental chamber was then set to run at a humidity level ranging from 3% relative humidity to 5% relative humidity, with the temperature variation from -5°C to 40°C.

In order to determine a control value, empty test tubes were placed in the environmental chamber for the sake of seeing the quantity of water that forms on the test tube and that does not arise from the substrate. All the water in the test tubes was then quantified using a pipette.

Following determination of the aforementioned control value, the steel, copper, aluminum, plastic, and glass wires were suspended on a holder in the chamber while the chamber was allowed to proceed through its nocturnal cycles. The more malleable wires were bent using pliers and suspended as shown in Figure 2, whereas the more brittle wires were held up using hooks. After 6 trials for each substrate, each lasting three complete 24-hour periods, the water yield was determined by subtracting the control constant from the water yield in each trial.

After using wires, the experiment was repeating with the use of water-filled tubing. The tubes were then sealed on both ends to prevent leakage. The tubes, made of steel, plastic, rubber, and glass, were all suspended with hooks. The same process was repeated once again, with 6 cycles for each tube, subtracting the constant value from the water yield each time.


Mean Water Yield (mL)
Aluminum / 0.0
Copper / 0.0
Plastic / 0.0
Steel / 0.0
Glass / 0.0
Control Value / 0.2

Initially, experimentation was conducted using wires of the varying sorts. However, as shown in Figure 3, the results were essentially inconclusive as no water yield was produced from the functionality of wires. A control value of approximately 0.2 mL was found after the wires were proven to be ineffectual.

The second experiment, which utilized water as a coolant to improve yield, produced more favorable results. Between all of the groups, the average water yield was 2.087 mL of water, with the control value accounted. The values obtained (Figure 4), while small, demonstrate the relative functionality of the system. Apparently, glass is the material that produced the highest amount of water, followed by rubber, plastic, and

Water Yield of Various Materials (mL)
Trial / Rubber / Plastic / Steel / Glass
1 / 3.3 / 1.4 / 1.3 / 3.5
2 / 2.1 / 2.1 / 1.7 / 3.1
3 / 2.4 / 0.9 / 0.8 / 3.4
4 / 1.9 / 1.7 / 0.9 / 3.5
5 / 2.1 / 1.2 / 1.2 / 2.9
6 / 2.6 / 1.6 / 0.8 / 4.0
Mean / 2.4 / 1.483 / 1.117 / 3.4

finally steel.

Obtaining simple numerical data is valid, but in order to analyze any trends or validity of the data, some statistical analysis was conducted. For the arrangement of data obtained, an ANOVA test was used to determine whether or not the means of the various groups were statistically different from one another and whether there was significance in the data.

The ANOVA test was conducted at α = 0.05, with the necessary critical value being 3.098. The null hypothesis stated that there would be no difference in the means, but after performing the test, the null hypothesis was rejected because the p-value obtained was almost close to zero, and thus significantly less than the value of α, and the F value was greater than the critical value.

ANOVA Summary Table
SS / DF / MS / F / p-value / Critical Value
Between Groups / 18.763 / 3 / 6.2544 / 35.7737 / 3.13E-8 / 3.098
Within Groups / 3.496 / 20 / 0.1748
Total / 22.259 / 23

After learning that there was a difference, a post-hoc Tukey test was conducted to determine where exactly the difference lay. It was found that the mean of rubber was statistically different from the mean of glass, plastic, and steel, that the mean of glass was different from all the others, but that the means of plastic and steel were not statistically different from one another.


/ Plastic / Steel / Glass


/ 2.400 / 1.483 / 1.117 / 3.400
S. H. (J/g°K) / 1.6 / 1.12 / 0.438 / 0.84

A correlation test was then conducted to determine whether these different values were affected by the specific heat of the substance. The mean values were set as y-values and the specific heats corresponding to each set as x-values. These values, shown in Figure 7, were used to determine a value for the Pearson coefficient, r.

For significance, the absolute value of r must be larger than the critical value. The calculated value of r based on the data was 0.301, whereas the critical value was 0.878, thus causing the null hypothesis that stated that there is no relationship, to be accepted rather than rejected.