In Vapor-compression refrigeration, a refrigerant is compressed releasing heat (typically outside a building) and then decompressed in another location, absorbing heat.

A residential air conditioner works on this principle with a compressor outside the home and a blower over the decompression coils to circulate cold air.

Would it be feasible to compress ahead of cooling demand, storing the compressed refrigerant in a high pressure container for use at a later time of the day, thus shifting the power consumption to times of the day where there is less electrical demand?

By feasibility, I'm thinking of questions like:

  • How much storage volume would one need for a typical household usage over an afternoon of running air conditioning given a container type that would be available cheaply, ex: propane canisters?
  • What would be the sources of efficiency loss added by something like this?
  • etc.
  • $\begingroup$ Storage of the liquid refrigerant would take a fairly large tank. The much bigger problem is how to store the decompressed gas afterwards. sCO2 might work though! $\endgroup$
    – Neil Belt
    Commented Sep 27, 2022 at 8:59

2 Answers 2


I happen to have a bottle of refrigerant in my house and another in my car (besides the stuff already being stored in my AC units).

While this is definitely more of a physics question, or perhaps a question for the HVAC guys on the Home Improvement SE... to generally answer your question, there are many factors for cooling a home, like size and location, for example. But let's say your average 2-ton window-unit AC pumps 24,000 BTU's per hour... and 4000 BTUs requires 1 gallon of refrigerant per minute.... you need 6 gallons per minute (being released through an evaporator coil) to match a window unit AC.

  • $\begingroup$ So 6 gal/min * lets say 2hrs of cooling = 720 gallons. Could that be reduced further by additional compression? For example, a quick search reveals that a typical A/C compressor is about 200-400 PSI, whereas, for example, the hydrogen tanks in a toyota mirai can store hydrogen at 10,000 PSI. Does pressure/volume work linearly like this such that going from 720 gallons at 400PSI would give you 28.8 gallons at 10,000 PSI? $\endgroup$
    – Gregable
    Commented Jul 26, 2016 at 18:19
  • $\begingroup$ That's liquid refridgerant... which is just run around in circles. Liquids are not really compressible. At 10,000 psi you might get 1% or 2% compression. So at 10,000 psi you need ~712 gallons. Or at 100 psi you would need 720 gallons of liquid refrigerant (for 2 hours use). $\endgroup$ Commented Jul 26, 2016 at 18:41
  • $\begingroup$ Ah, bummer. Just curious, but this doesn't sound very feasible. Thanks! $\endgroup$
    – Gregable
    Commented Jul 26, 2016 at 19:08

A BTU is a unit of heat, just like joules or calories. An air conditioner's capabilities are measured in cooling/time and $\pu{24,000 BTU}$ is colloquial used to mean $\pu{24,000 BTU/hour}$. The circulation requirements would be defined by

X gallons of refrigerant circulating per hour would yield Y BTU/h of cooling capacity

Apparently (I'm not an AC guy) that translates to ~$\pu{66,7 BTU/minute/gallon}$, or $\pu{4,000 BTU/hour/gallon}$.

If the main AC unit's cooling capacity is $\pu{24,000 BTU/hour}$, which comes down to about $\pu{7 kW}$ or $\pu{25 MJ/h}$ (that's astounding for a home appliance!), you need a certain amount of refrigerant in constant cycle.

If the assumption is $\pu{4,000 BTU/hour}$ per gallon of circulating refrigerant per minute, you would then need at least $6$ gallons of refrigerant to reach the 24,000 BTU/h. More is always a possibility, though.

P.S. I speak metric, so I rounded off enthusiastically in conversions.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.