Air-cars (powered by a pneumatic storage tank) have a bit of a problem with energy waste.
The ones I've seen on the net use multiple compressors to get the air to the 4500psi maximum pressure of the tank, cooling in between each compression stage. The reason of course is that as air is compressed,
its temperature rises which tries to expand the air: the upshot is that unless you drive away immediately, your tank will depressurize as it's contents cool to ambient. That's a fair bit of wasteage unless you opt to preheat your house hot water tank or something while refueling.
Also I postulate that the tanks are uninsulated for the same-but-opposite reason: while driving you'll be decreasing the pressure inside the tank, thus lowering the temperature from ambient and lowering the pressure even more: the lack of insulation allows the tank to rewarm up.
But bear in mind the typical usage pattern for these things: daily commutes and runs to the grocery store and back followed by an overnight recharge.
If you insulate the tank with an inch or two of styrofoam (or whatever) and toss some water into it you will waste much less energy...
The water absorbs the heat from the incoming pressurized air and, thanks to the insulation, will keep the air warm (and the pressure up) during the course of the day. Due to the totally awesome amount of heat that water can absorb, the air isn't at surface-of-the-sun temperatures and thus, a reasonably large amount of air can be pumped in.
As per the amount of water that should be used ? Optimum is such that if it was at ambient pressure, the amount of heat it had absorbed would be almost but not quite enough to boil it. (ie: if the tank breaks, the water won't flash into steam and void your warranty, so to speak)
Maybe somebody whose desk isn't covered with computer junk, flotsam and jetsam and various people's cats and can find their pencil, paper and little engineering formula handbook, can answer that one.