Gas line breaks you valve off the supply. a line on a AC unit breaks it will blow off until the pressure is gone.
From the Wikipedia article on refrigerants:
The critical point is the combination of temperature and pressure at which both the gas and liquid phases can exist. A refrigerant wouldn’t work well at or above its critical point because it the phase transition would stop happening.
In the spirit of the straightdope, I ran calculations with the OP’s configuration using the following refrigerants : Air, CO2, Ammonia, Propane and R12 (CFC). I used a cooling duty of 1919 BTU/hr (0.16 tons) - just enough to freeze 10 lb/hr of water at 80F.
The results are here : Imgur: The magic of the Internet
The designs are not optimized. The compressors need intercoolers for many of the cases (ignored) and the pressures can be tuned for better numbers. However, the results will illustrate a lot of points people have been making above.
Refrigerant selection is primarily a overall cost (over the lifetime) decision.
1> The capital costs (or the initial costs) are affected by
a> The cost of the refrigerant itself
b> The compressor (higher pressure - higher cost, hazardous or flammable fluid - higher cost, corrosive material - higher cost) etc.
c> Static equipment costs are primarily a function of size (how much refrigerant is needed) and pressure (also special metallurgy if needed)
d> Instrumentation / Controls get more expensive with hazardous / flammable fluids (sometimes exponentially)
e> Capital cost is also affected if there is a need for spares (that is the system is not reliable)
f> The cost of storing the inventory or flaring it when the system is down
2> Operating costs are effected by :
1> Power consumption
2> How many days of the year the system runs
3> How easy it is to get the refrigerant (at the right concentration)
4> Leakage rates of refrigerants
5> Maintenance which is linked to the corrosiveness of the refrigerant.
There are other factors too like regulatory ease and turndown which effects the bottomline.
The Einstein-Szilard refrigerator, see Einstein refrigerator - Wikipedia has no moving parts and no working fluid. The Wiki article has no discussion of the efficiency. They developed it in response to reports of the death of a Berlin family as a result of a leak in a refrigerator that used ammonia as its working fluid. It was patented but has not been used commercially, AFAIK.
I wouldn’t call them “extremely safe,” I’d use “mostly harmless,” since ozone does matter just a bit.