I'll break this down into a few smaller sections to allow for easier digestion. I don't expect everyone to read our whole report.
First off this was my senior design project in mechanical engineering at Purdue university. Our goal was to develop a cooler, not a refrigerator, that would be self sustaining relying only on a battery source and could be fabricated with as many local materials as possible. Initially we looking into ammonia refrigeration cycles using technology such as the "icy ball" (http://en.wikipedia.org/wiki/Icy_Ball). This would have proved difficult to replicate and prohibitively expensive. Evaporative cooling was hit upon as a potential solution.
Zeer pot operation:
A very clever device is the zeer pot, made well known by Mohammed Bah Abba who is able to produce the pots for less than 2 dollars American. They can significantly extend the shelf life of produce allowing farmers the ability to hold onto crops to receive better market prices, and individual families to store food longer. The device is simple with a glazed pot sitting concentrically inside of another larger porous (unglazed) pot and fine sand filling the gap. Water is added to the sand until it is completely saturated. Entropy is removed from the system by the water evaporating through the clay thereby cooling the inner pot. All of components can be found pretty much anywhere in the world.
The largest downfalls to this design are:
It must be tended to quite often, refilling water and such.
Approximately 2 liters per day are added to a large zeer pot. In some areas, it is more beneficial to drink the water than preserve food.
There is no proper lid nor insulation.
They only work in low humidity regions and rely on fickle winds to be effective.
The Evaporative Cooler (The Phyllis)
Our goals were to address these issues in an attempt to improve the device. Our thoughts were to capture the air after evaporating the water from the system and recondense the moisture so it can be reused. Also included would be a proper container that will both insulate from radiant heat (the ground, ambient air) and provide a reservoir that the cooler can use to replenish itself.
I would recommend looking at the poster presentation now to get an idea of the devices operation.
In order to control the airflow it was decided to push hot dry air through helical porous tubes buried in the walls of the container. In this model the air was pushed by two 120mm computer fans run by 12V battery. In the real application these would be replaced by solar powered fans, which are quite common in many places in Africa. However testing indoors made it impractical to use them in this iteration. The tubes were made of bend chicken wire covered in cloth. Again the focus here was on the ability to be replicated by providing very little if any materials. A template if you will that can be replicated with whatever is on hand.
The air would then travel underground and collect there to be pumped up every two days. This portion of the experiment was never fully tested due to reasons I will get into later.
The sand was kept moist by a reservoir of water at the base of the device and the natural capillary action of the sand was more than sufficient to pull the moisture up and saturate evenly.
The outer body was just a standard 10 gallon water cooler. It was chosen becuase it was insulated and had a proper lid that seals. The spout was replaced and used as the connection point for the pump and adding water to the reservoir.
Since the device only works well in hot, and low humidity areas , testing was essential. We were fortunate enough to find the biology building has large walk in incubators that are kept at a constant 38 degrees C (100F) and kept under 20% humidity. This eliminated many of the variables that would skew our data. We were limited by the accuracy of our sensors, being on a budget they were fairly cheap temperature gauges that measured both humidity and temperature. We’re relatively confident in their performance however. They all agreed within a small margin of error to each other.
The underground portion was intended to be simulated by a water bath that would keep a constant cooler temperature. We found that variation in temperature was reduced significantly in as little as 4 feet underground. There it will be approximately equal to the year round average. However since water baths are extremely expensive, the one we were able to find wasn’t able to cool any longer. It could heat water but that wasn’t useful to our application and eventually that portion was abandoned. The principle of the device would remain the same without the condensation portion.
Several parameters were changed from trial to trial isolating each variable (number of fans, number of coils, length of tubing).
We found that our design bottomed out at about 22 degrees C (71F) which is a pretty decent drop in temperature considering its just evaporating water. By the way the temperature measured was the air inside the food storage pot. This temperature was supported by some rough thermal predictions although systems like this are so complex that hand calculations really are not that useful. However it does show a trend of reaching the most energy that can be removed from evaporation of water. It suggests that the lowest temperature, under our parameters, is about 17C (62F). This temperature is far superior to ambient conditions for vegetables, however meat would need to be kept under 40F, similar conditions for vaccines.
We think that there is too much head loss in the piping, and reducing this would make the device more effective. In other words eliminate any bends and diameter changes to make it easier to push air through.
Again all this is explained much more thoroughly in the report and poster but I wanted to give a quick summary here as well. Please excuse any typos in them, I cannot seem to find the final version of the poster and report that had a graph's label corrected.
Anyway I thought I would post what my team had learned under our own experiments as I believe the idea of recapturing water and reusing it was a novel take on the device. We don’t have an accurate way of guessing how much water it uses per hour compared to the pot in pot, however in side by side tests it required no upkeep and each time there was more than enough water left in the reservoir to continue running.