February 11, 2000

TO WHOM IT MAY CONCERN:

I was asked to do a laboratory scale test of the effectiveness of the product watersavr as a retardant to evaporation of an exposed-to-the-air water surface. The following is a detailed account of the experiment and the results which were quite positive.

Apparatus: Two identical pyrex dishes were purchased in their boxes at a local store. Both were scoured with crocus cloth and a biodegradable laboratory detergent (Sparkleen). Washed with tap water 4 times and once with distilled water then filled with just over two litres of distilled water.

Procedure: The two dishes were placed on a bench at about 2.5 meters from an operating chemical fume hood. A black mark, made with a pen using water insoluble ink was made at approximately 1.65 inches (4.45 cm) above the bottom of each dish and a white adhesive paper strip applied on the outside surface behind the mark to increase visibility. The dishes were 2 inches deep and held about 3 L of water when completely full. It was found that the procedure of filling until the meniscus bulged over the edges was not satisfactory as in the watersavr covered one, the temperature rose, the water expanded and dripped over the dish's edge, taking most of the watersavr with it. The two dishes had a surface area of just under 1000 sq. cm. which decreased slightly with depth. Since the black mark on each of the dishes was done by eye, using a plastic ruler, filling to the mark did not give exactly the same amount of water in each dish. The uncovered dish held 2.2 L and the watersavr covered dish held 2.5 L at the mark. The dishes were placed side by side and a thermometer placed in each. A calculation showed that 0.2 mg of watersavr should produce a bilayer on the covered dish, but it was impossible to add such a small amount and enough was added so that the watersavr could be seen to spread. It is estimated that up to 10 monolayers of material was added. The initial spreading rate of visible particles was estimated to be more than 10 cm/s or more than 7 kph possibly as much as 10 kph. Smaller particles would be expected to spread faster. The dishes were undisturbed for nine days, that is 210.5 hours in a clean laboratory atmosphere at about 23 degrees Celcius and a relative humidity of about 35% except for the included two day weekend when the temperature was about 20 degrees Celcius and about 37% R.H. During all of the exposure time the covered dish water was about 1.5 degrees Celsius above that of the uncovered dish.

At the end of the test period the two dishes were filled from a 500 mL graduated cylinder again to the marks and the amount of distilled water added noted.

Results: The uncovered dish required 1.362 L and the covered 1.150 L to fill their respective marks. The difference of 212 mL lost during the 210.5 hours of exposure was exactly 1 mL per hour per 1000 sq. cm. within the limits of the expected experimental error. In terms of percentage of water lost by the two dishes, the uncovered dish lost 61.9% and the covered dish 46%, or the uncovered dish lost almost one and one half times as much water as the covered dish. In terms of yearly loss, 88,255 mL (88.2 L) would be saved for every 1000 sq. cm. or about 88 cm (almost 3 ft.) of water from a reservoir.

Comments: The laboratory scale experiment must be viewed as "proof of concept" only since field experience on large, deep bodies of water should be different to some extent. It should be pointed out that the covered dish water showed at least 1 degree Celcius higher temperature - a proof of the retarding effect on evaporation of the watersavr. Since the vapour pressure of water is higher at the higher temperature (about one mm Hg per degree Celcius in the range of temperatures of the test) the effectiveness of the watersavr must be greater by a difficult to calculate amount than the values calculated in Results above. Again, since the cubic expansion of water has not been used and the density varies by about 0.00022 g/mL per degree Celcius in the temperature range of the experiment, filling to the marks with cooler distilled water of the two dishes would have increased the difference in water added if the difference in evaporated water had been measured by weighing. Finally, it is estimated that these small increase in the ability of watersavr to save more water may perhaps have raised these results by only as much as 10%. In large, deep bodies of water, these minor effects would not occur, though surface temperatures in a covered reservoir under still air conditions would be expected to be higher, but not so much as in these shallow dishes.

Signed,
Robert N. O'Brien,
Professor Emeritus