The bottom line is that the air lost required a miniscule amount of heat
loss to cool to its temperature when lost, and that what replaces it
will require a miniscule amount of heat loss to be cooled itself.

So far as the milk spoiling, that is the ultimate point, isn't it?
either the unit does its job reasonable efficiently, or it doesn't.  the
reason I have a fridge is to keep the milk, eggs, and other food items
cold and fresh.  I use a cooler for beer.

If you are really worried about that air, just make sure to (1) pre cool
the stuff you put in, and (2) keep the unit fairly full (while allowing
for reasonable air circulation to prevent "dead" zones).

On Tue, Sep 8, 2009 at 12:51 PM , Steve Williams wrote:

> At 09:56 AM 9/8/2009, Rocket J Squirrel wrote:
>> It takes a pretty long exposure to warm air to raise the temp of a
>> can of beer even a couple degrees.
>
> Oh, right, well, I understand that the stuff in the fridge doesn't
> warm up much when I open the door.  That's not what I was concerned
> about.  I understand I'm making my fridge more efficient by storing
> many more BTUs of cold per cc in my 3.5-litre water jug than if I
> just left air in there.
>
> What I'm concerned about is the loss of the cold AIR that inevitably
> fills the space between the bananas and the bottle of salad dressing.
>
> When you open the door of a stock fridge to cool off your toes, how
> many grams of propane or millwatts of electricity does it take to
> cool the warm air that replaced the cold air that flowed out?  (As
> opposed to just leaving that cold air inside, as the top-opening
> fridge mostly does.)
>
> I have done that calculation, but, as you say ...
>
>> ... we hear the reefer's compressor kick in, but it is responding
>> only to the temperature of the air ...
>
> That's my whole concern.  I'm not worried about the milk spoiling.