, ,

I’d read many positive testimonials in relation to the legendary pre-electric cold-store unit, the Coolgardie safe… but I still wondered how well did it really work? When curiosity finally got the better of me, the Eldorado Museum Association came to my aid with the genuine article.


Coolgardie safe, courtesy Eldorado Museum, Victoria

In my first post of 2017, I’m going to momentarily diverge from the 1850s gold rush. Meanwhile, I hear you asking, ‘What is the Coolgardie safe?’ The short answer is ‘a meat safe’ — but there’s more to it than the Coolgardie safe being a ventilated, vermin-proof box in which to store perishable food.

The Coolgardie safe was invented in the 1890s on the goldfields of Coolgardie, Western Australia, by Arthur McCormick. McCormick had observed that a wet bag placed over a bottle cooled its contents, and that if this bottle was placed in a breeze, the bag would dry out more quickly, but the bottle would get colder. What he was observing was the basic principle of heat transfer that occurs during the process of evaporation: that as it transforms from a liquid to gaseous state, water consumes energy in the form of heat, taken from its surroundings. [1]

Designed to take advantage of the cooling effects of heat transfer from evaporation, the Coolgardie safe was a common household item in Australia until the mid-twentieth century, initially vying with iceboxes until finally overtaken by kerosene and electrical fridges. I’d read many descriptions and testimonials regarding the Coolgardie safe, but I still wondered if it worked, and if so — how well?

I knew that the Eldorado Museum had one in its collection [2], which, until recently, had languished ‘out the back’ until local Howard Phillips gave it a fresh jacket of hessian so that it could go on display. Once on display, members of the Ross family from Wangaratta recognised it as a model that had been manufactured by their family business in Wangaratta (albeit more than half a century ago), and so they donated a set of brass name plates to attach to it. It was my good fortune that the Eldorado Museum Association agreed to let me borrow this safe for a living history trial.


Manufacturing plate on the Coolgardie safe at the Eldorado Museum.

This particular Coolgardie safe is of a common type: a rectangular metal frame, which supports hessian sides, wired on. It has a simple hinged door on the front, and one internal shelf. The top of the frame has a galvanised sheet-metal tray (ie: a reservoir), which is filled with water. Strips of flannel are hung from the tray to contact with the hessian sides, which keeps them damp through a process of capillary siphoning (wicking). When a breeze comes, it passes through the wet hessian and evaporates the water. This cools the air inside the safe, and in turn, cools the food stored inside. The drier the air is, the greater the rate of evaporation, and the cooler the safe. The feet of the safe sit in a tray of water (also on legs), which acts as a moat to keep ants at bay, and collect water dripping from the hessian. Sometimes these trays had a tap to drain the water, but this model has only a simple spout. There is also a central vent in the bottom of the safe which vents through a little chimney passing through the middle of the reservoir at the top — and I’ve seen other Coolgardies with this feature.


Another Coolgardie safe complete with a lid over the water reservoir. This safe has had wire mesh added at a later date (probably for decorative effect) where the hessian panels would have sat. (Photo swiped a long time ago from a seller on eBay!)

Day 1, 9 February 2017

I put the safe under my back verandah and filled the reservoir to the very top, which really kicked-off the wicking process: water could be seen wicking from the flannel down and across the hessian panels in a matter of seconds. Initially the strips of flannel I used were only half the width of each hessian panel. To increase the amount of water wicking through the hessian, I doubled the flannels until they were virtually equal to the hessian panels. This optimised the wicking, and the safe began dripping continually from the bottom, as water ran off. Immediately, I could see why it would be necessary to have a drip-tray — simply to conserve and recycle water. (Ideally you would find an economic balance between enough wicking to make the safe functional, while preventing too speedy a loss of water from the reservoir.)


The water reservoir at the top is filled, and strips of flannel are draped over the sides to contact with the hessian panels.

After the reservoir was filled to the brim and the sides were wet, at 12.56pm, it was 30.5ºC in the shade of the verandah and 28.6ºC inside the safe, rendering a difference of 1.9ºC in temperature. (The thermometer I used was highly sensitive and constantly went up and down as much as 0.2ºC, and what’s more it read different temperatures under different places around the verandah, so my readings are a little rough.) I continued to monitor the temperatures, with these results (contrasting the temperature under the verandah with the temperature inside the Coolgardie safe, showing the difference between the two):

1.23pm — 29.6ºC vs 28ºC = 1.6ºC
1.32pm — 29.2ºC vs 26.4ºC = 2.8ºC
2.15pm — 32.3ºC vs 29.6ºC = 2.7ºC
2.42pm — 28.9ºC vs 25.6ºC = 3.3ºC

Perhaps the temperature would have continued to drop, but I decided to add a large pedestal fan to see if this would increase the rate of evaporation, and thus decrease the temperature. The short answer is, yes, it did.

3pm — 31.4ºC vs 26.4ºC = 5ºC
3.22pm — 33.6ºC vs 28.5ºC = 5.1 (and this temperature steadily fell…)
3.39pm — temperature inside dropping to 25.2ºC
3.42pm — 32.6ºC vs 24.7ºC (and my guess is, still dropping) = 7.9ºC

So I could say that the Coolgardie safe, when wicking steadily, ran from a bit over 2.5ºC to 3ºC cooler than the outside temperature, and between 5ºC to perhaps as much as 8ºC cooler when there’s a decent (albeit artificial) breeze.

Frankly, I don’t consider these to be amazing results. However, this was all without the drip tray beneath the safe. The drip-tray that came with the safe was rusted-through, so I hadn’t used it; but now I was beginning to wonder whether the tray may have also provided some micro-climatic benefits.

This question is something I’ve decided to pursue on the ‘morrow. As I write this, I’m about the pull a beer, chilled to 4ºC, from my fridge.

Day 2, 10 February 2017

I didn’t get started until later in the day, and it was a scorcher, topping a bit over 39ºC! This time I put the drip tray under the Coolgardie safe, and filled both it and the reservoir at the top. I didn’t use the fan at all. The results speak for themselves:

At 3.15pm it was 39.1ºC degrees in the sun, 36.9ºC under the verandah, and immediately 31.4ºC in the Coolgardie. This dropped to 30.3ºC within the next ten minutes.

Here are some readings:

3.15pm — 36.9ºC vs 30.3ºC = 6.6ºC
3.45pm — 35ºC vs 26.1ºC = 8.9ºC
4pm — 34ºC vs 27.8ºC = 6.2ºC

Now I could say that when fully set up, the safe is at least 6ºC cooler and maybe as much as 8-9ºC cooler than the shade of the verandah. These are the best results I could achieve on a very, very hot day, without the aid of a fan. I admit this wasn’t the most scientific trial, and maybe there are other ways to achieve better results. If you have any ideas, please let me know.


Thanks to my sponsors, the Eldorado Museum. This is the second living history trial in which I have been ‘enabled’ by members of the Eldorado Museum Association and other friends from Eldorado. The first trial was the manufacture and use of a replica gold cradle based on one held in the Museum’s collection. You can read about that here.


[1] In my history of the Coolgardie safe, you may recognise information swiped from the Powerhouse Museum and Museum of Victoria.
[2] Indeed, there are very few domestic household items from the mid to late nineteenth century that the Eldorado Museum doesn’t have in its collection!