The Yurt Cooler is the creation of two electrical engineers (Luke Allen and Chris Gervang) who decided to engineer the best swamp cooler possible for Burning Man 2015. This page documents our calculations, tests, and what we learned.
Test of the final version at Burning Man 2015
This test occurred around 5pm. Outside temperature in sunlight was 94F, in shade was 87F. After about 10 minutes of operation, the Yurt Cooler outlet temperature was 50F, and the temperature inside the hexayurt was 76F and dropping. (Not shown in the video: the temperature eventually stabilized at 74F.)
The Yurt Cooler proved to be robust, and worked reliably all week. See below for how we designed it.
Background: Evaporative Cooler Principles of Operation
For fancy physics reasons, liquids absorb heat when they evaporate. (E.g. one gallon of water absorbs about 8700 BTU when it evaporates). Therefore, you can cool your house by simply evaporating as much liquid as possible inside it.
A conventional air conditioner does exactly that: it evaporates liquid inside a metal pipe in your house, absorbing heat. It pumps the resulting vapor outside, then compresses it back into a liquid (releasing the heat that was absorbed). It repeats that cycle forever. This works amazingly well, but the compressor needs a lot of energy. It can't be powered by a reasonably-sized battery.
An evaporative cooler (a.k.a. "swamp cooler") skips the compressor. It just evaporates water inside your house, and lets the vapor escape. This means it consumes very little energy, though it does consume water. And it only works well in dry air, where it's easy to evaporate large quantities of water quickly. This means it works great at Burning Man.
Efficiency Optimization
The Yurt Cooler evaporates water by simply pushing dry air, from the outdoors, through moist humidifier filters, and into your house. The evaporating water cools the air, resulting in a flow of cold air into your house.
One gallon of 90F water absorbs 8700 BTU when it evaporates, so the BTU per hour absorbed by a swamp cooler is just
BTU absorbed/hr = 8700 BTU/gal water x (gallons of water evaporated per hour)
However, the swamp cooler must bring in dry outside air, which brings in heat. So, to calculate cooling power, we must subtract the heat added to your house by the hot air. (In other words, the evaporating water must first cool the incoming air down to your indoor temperature, just to break even. From there, any further evaporation actually cools your house.) So to maximize cooling, it's vital to evaporate as much water as possible per pound of air we bring in. So, we want to get the Yurt Cooler's outlet air as close to 100% relative humidity as possible.
To achieve this, we tested the Yurt Cooler using a humidity sensor. After optimizing air flow rate, as discussed below, we added humidifier filters in series until we got close to 100% relative humidity. We initially thought we might have to force air through a tortuous path of many filters, but we found that, for our flow rate, 3 humidifier filters was sufficient to get above 90% humidity.
Airflow Optimization
Our rate of evaporating water is limited by the amount of air we can push through the humidifier filters. If we assume that our humidifier filters are doing their job, and evaporating as much water as possible into each cubic foot of air, then our cooling power is directly proportional to how many cubic feet of air we can force through the filters. We used an anemometer to optimize air flow with a series of tests. These were our big lessons learned from our tests:
Computer fans can't produce enough pressure to efficiently push air through humidifier filters. The back pressure from the filters reduces them to a tiny fraction of their advertised cubic feet per minute (CFM) airflow rate. Boat ventilation blowers are designed to push against back pressure and do much better. (More precisely: a given fan can be characterized by a "pump performance curve" that plots CFM flow vs. psi back pressure. For this application, we would like a fan that maintains high flow when there is back pressure. However, manufacturers of cheap fans generally don't supply a performance curve, so we just bought a bunch of fans and tested them. Boat ventilation blowers were the winner.)
Any sort of constriction in the airflow path (such as a small diameter outlet hole, or a ventilation duct) will resist flow much more than the humidifier filters do. For example, adding just a short section of ventilation duct reduced the flow to around 30% of the original. An outlet hole that was the same diameter as the blower dropped the flow rate to around 50% of the original.
We initially tried inlet filters with the same diameter as the blower; those also caused a big drop in airflow.
From the above tests, we concluded the following: To maximize air flow, the entire flow path must be as wide and open as possible. Ideally much wider than the blower diameter. The inlet filter and the outlet hole should be as big as possible. Ventilation ducts should never be used.
The final Yurt Cooler design on this page incorporates those lessons. It achieves around 60CFM airflow, which is much greater than other DIY swamp cooler designs.
Hexayurt-specific optimizations
We programmed this Python script to calculate heat flow in and out of a hexayurt, including the effects of the Yurt Cooler. These were our conclusions:
In daytime, the main heat flow into a cooled yurt is the heat conducted through the walls and the taped seams. Heat added by a couple human bodies is actually negligible compared to heat added through the walls by the sun.
Heat conduction to the ground is also negligible, when the ground is dry. The amount of heat that can conduct through a foot of dry earth is tiny compared to the heat conducted through the walls. (On hexayurt forums, you sometimes read claims that the ground will help cool your yurt. That is not true in places with dry soil such as Burning Man.)
So, the only effective ways to reduce heat flow into a yurt are to put it in the shade, use thicker insulation panels, and ensure that you don't have large gaps in the insulation panels at the taped seams.