The Great Canadian Refrigerator Experiment

Ok, so not everything around here has to be serious, so when Rob posted on Facebook a definition of the "Canadian Refrigerator" we were motivated to perform some good (or quite possibly bad) science to confirm or deny this hypothesis.

For your reference, the definition states:

Canadian Refrigerator:
A bank or pile of snow during the colder months of winter, where food or beverage products (namely soda and beer) can be stored. Great for parties.

"Are we out of Molson Canadian?"
"There's plenty in the Canadian refrigerator."

So, since I live in the coldest part of provincial Canada at the moment (semi-northern Saskatchewan) I thought I'd give it a try.

The weather today was a brisk -37 degrees Celsius (which is essentially the same in Fahrenheit)

The theory we're testing is that the snowbank keeps your drink cool but unfrozen due to the insulating properties of snow. However, since beer isn't thermogenic (like a person) I suspected it would freeze nearly as readily IN the snow as out of it (although perhaps not as quickly). However, the definition implies that you can store the soda or beer outside in the snowbank for some indefinite period of time.

On to, the experimental criteria:

A beverage, namely a Smirnoff Twisted, Green Apple flavor (no beer was harmed during this experiment).

Cold outside temperatures:

A snowbank in which to store the beverage in question.
The snowbank should be reasonably deep and wide as to provide the required insulation or cooling but not so deep and wide as to be impractical for storage. The resulting snowbank was artificially constructed and approximately 3-4 feet deep.

The beverage would be stored at a practical depth (approximately 6-8 inches under the snow)
The time spent photographing the subject would be minimal (10 seconds or less) to prevent the beverage from freezing from exposure.

Experimental observations:

At 5 mins, 10 mins and 20 mins the beverage demonstrated no freezing effects although I noticed what might have been an increase in viscosity of the liquid contents.

At 40 mins, the beverage looked mostly frozen:

At 50 mins, the beverage looked totally frozen and was partially confirmed by removing the cap and sticking a probe into the frozen contents

The experiment was concluded at 50 mins for fear of the beverage exploding and leaving bits of glass and slush in the snow that would need to be cleaned up.

After some time the beverage thawed and started to push delicious looking slushy through the neck:

The experiment's conclusion:

Conclusion
So, while the Canadian Refrigerator may work in warmer temperatures, it clearly fails in -37C and since no reliable standards exist to determine the maximum time before freezing, it probably isn't worth using as a temporary cooling method either. It seems reasonable to expect that since the beverage isn't thermogenic and is generally already cool when purchased, it doesn't have enough thermal mass or adequate internal temperature to survive for long (without freezing solid) in anything but modestly cold temperatures. However, if a reliable method could be determined or predicted to determine the eventual freezing time, it would probably make a great temporary cooling system (to get the beer as close to freezing as possible).