Solar energy to chill water - store energy

[ThaTermeneter]

I am hoping someone can help me with the mathematics of my idea. Maybe some HVAC or engineering guys could chime in?

I would like to have a small air conditioned room, powered by solar energy. The problem is of course that the sunlight is only available a limited amount of time each day so the solar panels need to collect extra energy so the system can operate into the evening when it is still hot. Battery banks are expensive, heavy, usually require maintenance and vent hydrogen gas, and their capacity slowly decreases becoming less and less useful - eventually needing replacement.

So my idea is, size the solar array so that it can start and run the air conditioner constantly without the need for chemical batteries. The air conditioner would be modified so the original "evaporator" would be relocated - submerged in an insulated reservoir of water/antifreze mix. Then there would be a seperate pair of radiators plubmed together, one in the reservoir and one outside to transfer the heat.

The idea is that the air conditioner runs more than it would in it's original configuration. This extra cooling time would allow the coolant reservoir to be chilled as cold as you could get it, and when the sunlight is done for the day you still have a bunch of cold water that can be circulated through the radiator with convection, or a small pump that would require a much smaller battery than needed to run the actual air conditioner.

What I want is someone that can help me figure out the math to size the reservoir. Say a 4000 BTU window unit could keep the room sufficiently cool during normal operation, how would I upsize the unit and determine the required reservoir capacity for x additional hours of cooling. I want to determine the feasability of this idea without actually building it.

[AngryOnion]

Look into a heat pump system.
You could use your back yard as a condenser of sorts.

[Mantikore]

Assuming you can consume the stored energy at a constant rate both when its being accumulated and depleted, and assuming its more or less an ideally insulated system (I probably dont need to say theyll be lots of energy leaking everywhere), cant you just calculate the amount of energy required per unit time with the "charging" phase, and then multiply it appropriately to get the energy required for use after-hours?

Or I could be looking at the system differently. Would the capacity of 4000 BTU's be an irrelevant fact, since energy isnt really stored at this point anyway??

[duncklee5]

These type of scientific experiments are very useful and gave such a nice info about this new experiment your information inspired me allot.