Carl Johnson at
http://mb-soft.com/public3/electzzz.html seems to have an effective and simple way of generating electricity, his invention basically uses low-pressure air/water to push a large piston. The forces involved can be large, for example if wind is used at 800Pa (80kg/m^2) you can get 400W if a 1m^2 piston takes 2 seconds to travel 1 metre. The machine can work off wind, flowing water, heated air or any other low-pressure fluid or gas.
The most interesting possibility he talks about is that his machine can very effectively capture solar energy, which is about 5kw/m^2 on a sunny day. He suggests building a greenhouse to capture the heat, which can then be pumped through his machine, and after all the losses are considered the efficiency is about 66%. He says he has built and tested a small version and it worked as expected.
The only problem is that if you're using low-pressure air you need huge amounts of it to make a decent amount of electricity, he says about 3m^3/second for 2.4kW. I'm having alot of trouble finding out how much energy you get from say 1m^3 of pressurised air expanding to 1.1m^3 at atmospheric pressure, hopefully someone who know their physics can help me here. Im guessing it's approximately P x DeltaV / 2? If so the machine seems to be weaker than Carl suggests, but he's a nuclear physicist and he obviously knows his maths better than I do.
He explains how his machine works but he hasn't built a full-scale version of his invention yet or explained exactly how you would pump so much air through the machine, I've drawn a plan of what I think might work.
Any comments would be welcome, e.g. how to make it more efficient or how to properly calculate the energy generated.
Operation (see drawing):
1) The 400kg piston (1m2) lifts 1m, drawing in 1m3 of air.
2) The 400kg piston drops, pumping the air through the greenhouse (H, say volume of 10m3), lifting the 395kg piston (with area 1m2) by 1m
3) the air is heated by the greenhouse (say by 10 C), expands (by say 0.07m3), and lifts the 395 kg piston another 7cm
4) The 395kg piston drops and the process repeats
Using P1V1/T1 = P2V2/T2 it seems that a temperature difference of 20 C will produce the 0.07 m3 increase in volume.
Total energy gain is -400x10x1 + 395x10x1.07 = 226J or 226W if it happens once/second. Does anyone know if this is useful/accurate?
It's simple to design/understand/build and produces the energy continuously, day and night, as the heat from the greenhouse can easily be stored in thermal mass.