This forum is about wrong numbers in science, politics and the media. It respects good science and good English.
Ricardo showed many years ago that water injection would improve fuel economy on petrol engines. It was tried out in practice on a municipal bus fleet (Bristol?) and shown to produce fuel savings of the order of 10 to 20 %. However the equipment was not particularly reliable and the extra costs of installation and maintenance pretty much ate up the savings so the idea was abandoned.
Its been revived at various intervals and always shown to work to some degree but never really well enough or consistently enough to to be properly and professionally engineered into a vehicle. Nearest thing to professional deployment was in big piston aero-engines as methanol-water mix injection to raise peak power. The large degree of supercharge employed on such engines leads the common assumption that the gain is due only to the charge cooling effect giving better cylinder filling. Clearly not something of substantial import with a normal road vehicle.
I'm somewhat unconvinced wondering how the large change in volume from turning water into steam, albeit at the cost of considerable input energy to drive the change of state, compares with the conventional heat the air and combustion products operation of the standard petrol engine. The question is whether there could be a net power gain by burning petrol to produce steam, sort of internal flash boiler system, than to produce hot air.
What doesn't help is that the ordinary, throttle controlled, internal combustion engine always settles at the least efficient operating point able to generate the required power (as torque into load). This is a major reason why automatic transmission cars fitted with torque converters show worse economy than fixed gearboxes. The torque multiplication in the converter allowing the engine to settle at a less efficient operating point whilst the transmission ensures that there is still enough torque available. Clearly there will be little benefit from an effective water injection system if it similarly shifts the part throttle operating point to a less efficient point on the engine load line curve. Where water injection has been shown to be beneficial the normal operating point has been towards the maximum power/maximum torque part of the engine load line.
To a first approximation the efficiency of any particular IC engine is related to the peak combustion pressure. Hence a variable expansion ratio engine, such as Professor Timoneys modification of the Tilling Stevens opposed piston 2 stroke diesel engine, has the best chance of usefully exploiting water injection. The Timoney system is elegant in that the two crankshafts are mounted in eccentric bearing housings interconnected by a simple spring and damper arrangement so the motor pretty much always operates at the same peak combustion pressure. At low powers the eccentrics swivel to lengthen the stroke and hence increase the expansion ratio with obvious gains to fuel efficiency whilst at high powers the stroke shortens and the engine runs less efficiently but at greater BMEP. One of the major proposed advantages of the system was that the engine could be sized to run efficiently at modest load, cruising, powers yet still have high peak output capability when needed. Conventional engines tend to be sized for peak power demands and hence be less efficient than they need be at cruise loads.
It might be interesting to create a modern version of the Timoney engine exploiting electronic engine management. Besides water injection, variable valve timing via sleeve valves should prove beneficial as would some sort of constantly variable ration transmission system arranged to operate in close to constant torque mode.
Ooops. Edit induced error needs correcting on my previous post. Plus a fe additional points.
The Roots / Tilling Stevens engine used as a basis for Professor Timoneys work uses rocking levers to transmit the motion of the opposed pistons to the single crankshaft mounted centrally below the cylinders. Its the shafts carrying the rocking levers which are mounted in eccentrics not the crankshaft as stated.
A disadvantage of this configuration is that it needs two connecting rods per piston and the rocking levers so there much more mechanical inertia than with a conventional engine. I guess that modern design and materials could reduce this problem but, inevitably, the maximum rate of rpm change will be less than with a conventional engine. Whether that is actually important under load is another matter especially as the power output changes needed for acceleration can be as much due to shifting the operating point of the engine, which is a fast process, as due to rpm changes as with a conventional motor. Proper exploitation of this characteristic is difficult with conventional transmissions.
This engine was a two stroke, externally blown, uniflow diesel but there is no reason why the configuration couldn't be used in a petrol motor whether two or four stroke. Given the detonation suppressing properties of water injection it could probably be operated at a higher compression ratio than a conventional petrol motor too.
The main interest in water injection or FWE (Fuel water emulsions) is for heavy fuel oils which need to be heated to help with atomisation. The water flashes to steam and provides better atomisation and combustion with reduced pollution.
One of the main benefits is the reduction of NOX which is reduced due to the cooler temperatures during combustion. (as I understand it)