WRT to Tata: ... the biggest reason that the assorted "experts", including a fair number of PhDs have looked like total fools and jackasses on this one (including moi, to a lesser extent) is that they naively "assumed" that the compressed air was being simply "expanded" instead of being used as a "virtual fuel".
Simple expansion would be pretty dumb and essentially non-functional. A little knowledge is a dangerous thing, as they say -- but Tata was also complicit in holding back some relevant info early-on - as it turns out in retrospect. Instead of simple expansion, the MiniCAT runs a normal 4-stroke piston engine that uses a modified "virtual fuel cycle" of exceptional efficiency. IOW the compressed air should be considered as liquid fuel operating at 100% Carnot (when in the engine, but not the total cycle as the losses are all in the compression). As with a diesel, outside air is drawn into the compression chamber and compressed to a very high ratio: twenty times or more to ~20 bar (290 psi). This air reaches 400°C and, at that point, the "virtual fuel" (air from the storage tank which although technically not liquid is "virtually" liquid) is injected into the "combustion" chamber, just as with the diesel. Here the fuel actually "explodes" with a large shock wave ("jerk" perhaps) not unlike real combustion but in a non-thermodynamic yet violent fashion, giving about one-fourth the normal amount of net expansion and torque as if it were "real" diesel fuel. It is actually only ~ 1/16 as energetic but the Carnot is essentially four times better. The exhaust is one fourth or less the temp of fossil fuel, in degrees K, but the shock wave is not diminished as much. The exhaust is about 250 K instead of 1000 K, so the exhaust can actually be used as air-conditioning - yet in the cylinder, the velocity of expansion is very high and fuel economy is therefore rather decent on a "per unit of mass basis" (not on unit of volume basis). Who cares if this cycle requires large storage tanks (except for the danger in a collision), since each fillup is one tenth the cost of gasoline i.e. about 20 KWhr of grid power ($2 bucks in many places). Pretty cool, kewl, or cuil ... I'd say... Jones