Hi
Just thinking about the real ice idea. Instead of using electric pumps to push water into the surface of polar ice, why not use carnot engines? They can work independently, in very remote regions of sea ice.
The principle is pretty simple. You'd have a high pressure, fixed volume chamber under ice, exposed to sea water. Above the ice is an expanding, low pressure chamber. Between the two is an upward-flowing pipe, and a downwards-flowing pipe - each with a non return valve. The up pipe has a turbine or piston engine in, which powers a water pump.
At the start of the carnot cycle, the high pressure tank is full of working fluid (eg chlorine, ammomia, DME) at atmospheric pressure. This boils, due to contact with warm seawater.
The high pressure tank then empties, turning the turbine and powering the pump. The low pressure tank then slowly fills up with low pressure exhaust gas.
When the high pressure tank is empty, (almost) all the working fluid is in the low pressure chamber, as a gas. This then cools, condenses and falls down into the high pressure chamber. The cycle then starts again.
Does this work? Is it better than an electric pump?
Andrew Lockley