Evaporating water – The Coolgardie safe
The classic of this technology is the Coolgardie safe, invented and used in the hot dry environment of the WA goldfields. It is usually a wooden frame with a hessian or similar fabric wrapped around it, with a water reservoir on the top and sometimes a catch-basin at the bottom. Water is placed in the top, soaks into the fabric and then moves down the outside of the safe by capillary action to drain into the catch-basin and on the way some evaporates, cooling the inside of the safe. That at least is the theory, but how does it translate into practice?
To start with I did a bit of research to see what information was around, one book which was very helpful was “New Australian Home Carpentry Illustrated” by Alex Smith, published by Colourgravure. Is not dated but looks to be early post World War II and I still see it second hand on a regular basis. There is a design on page 65 that, while it was too big for me, I used as a base for my design.
The material of construction that I used for the framework is 19mm x 42mm DAR pine. I used this timber because it was cheap (less than $1 per linear metre), available, light and easy to work with. The frame whole frame stands 1.5 metres tall which, as luck would have it, happens to be as high as the original four lengths of pine that I bought, funny how designs work out like that. The pine is not treated in any way and so is subject to attack by the water, which meant I had to apply a couple of coats of paint before putting it into service.
To start the frame I connected three cross pieces, top, centre and about 100mm up from the ground to each set of uprights with wiggle nails, these are cheap and nasty and if I had known how much work the rest of the safe would be I would have used dowels and glue, although at that stage I didn’t have a dowelling jig, but I do now! This gave me two sides which had to be joined together to make the carcass of the Coolgardie safe, but the question was how wide should it be? I decided to see what containers were around that could be pressed into service as a reservoir and catch-basin. Again it had to be light, cheap and readily available. My initial thought was the mind bogglingly useful pet litter tray, but in this case it proved too shallow, I wanted something with more capacity. I settled on two 360mm x 300mm x 120mm deep plastic basin (10-12 litre capacity) that retailed at the Reject Shop for the princely sum of $2.50 each, thus the distance between the uprights became 360mm, less about 15mm on each side so the overhanging ledge of the bowl would sit on it, thus providing a support. The lower catch basin would sit in the space formed by the lowest cross pieces if the over hanging corners were cut off.
So the next trick was to form up the carcase by installing a top, centre and lower rail on the back and a top and bottom rail on the front (to allow for the door), this I did by drilling and screwing through the side of the upright and into the end grain of the rail. Two screws on each side of each rail and the carcase started to look like a piece of furniture instead of a bunch of randomly assembled pieces of wood.
I needed to insert some shelves for the produce to stand on, my old friend the Aussie Home Carpentry Illustrated suggested using cement sheeting but due to lack of availability at the time and concerns for weight (I wanted the safe to be transportable) I constructed timber shelves. I suspect that the book was probably right in that the shelves would become wet and cool due to evaporation also, you can’t win ‘em all! I screwed in some supports to the sides of the uprights and then screwed pieces of the 19mm x 42mm pine to the supports to form an open grid for use as a shelf. Before the lowest shelf went in, the lower catch basin had to be fitted, because there was not enough clearance between the bottom shelf and the bottom rail to allow the basin to be slid in and out.
The door was a simple construction of two uprights with a top centre and bottom rail wiggle nailed to them, I didn’t put in any diagonal bracing and it hasn’t dropped……………..yet. This I fixed to the front of the safe with screws and some mild steel hinges (I considered brass but they were a bit expensive) so I had to ensure that they were well painted to prevent rust.
With the carcase finished and the basins in place it was time to fit the fabric cover. Again, simple and cheap was the rule so I bought a couple of hessian sacks that are used for dog bed covers, they were only $2 a piece at a local pet barn. They are sewn up the edges and I decide to fit them length ways down the safe, to allow for an unobstructed water flow, so I had to undo all the stitching which was quite tedious but the thread could be (and was) used to tie tomato plants to stakes etc. Each sack went the full length of the safe with enough left over to overhang into the upper and lower basin and was wide enough to neatly cover two sides, so I needed two sacks.
I attached the sacks with dome headed drawing pins which gives a nice effect. The first sack covered the left hand side and the back, the second covered the right hand side and the door, the idea being that the drawing pins were fixed to the leading edge of the door so that it could be opened. When the door is opened wide, the overhang into the top and bottom basin has to be tucked in again, but it was either that or arrange some form of flap to allow water to pass from the top basin onto the hessian door cover.
Something to be aware of is that the sack which was attached to door acted as a very efficient wick, while the other one, initially at least, did not seem all that interested in soaking up the water regardless of my threats, begging or bad language. I suggest that if you try this, soak both sacks in water first, perhaps with a little detergent to help the water “wet out” the fibres. After some use, both seem to be OK now.
So….how does it work?
This depends on such variables as ambient temperature, humidity and air movement and I suspect that while a Sydney summer is no hotter than the WA goldfields it is somewhat more humid. If the top basin is filled, it will empty into the bottom basin in about 12 to 18 hours and will result in the material inside being a bit cooler than the surrounding air, but I have not measured by how much. Much of this water is caught in the botton catch-basin and can be recycled.
The cooler did not work in the shady garage, due to the lack of air movement and it is now behind the rear window of the garage, I have intentions of putting a covering of shadecloth over the area to cut down on sun but allow some air movement. I must admit to being a bit disappointed with the cooler’s performance. Improvements to think about could include –
Retrofitting those compressed cement shelves.
Wrapping another layer of hessian around the cooler.
Installation of the above shadecloth.
The installation of a tap in the bottom basin to facilitate water recycling.
It may be that I have just not found the right place for it yet. One initial thought was that the cooler was too tall and the water would evaporate before it got to the end of the cooler and into the bottom catch basin, but this has not proved to be the case.
