🚀 go-pugleaf

O.K.
Using reference material within arm range.
osmotic pressure of a salt solution of 1000 mg/l = 78.6 kPa
therefore 35,000 mg/l approx. = 2751 kPa (not quite true for electrolytes but it
will do)
= 280 m water head
Density fresh water = 1 g/cm3
Density sea water = 1.025 g/cm3
 Column of water required to obtain pressure differential of 280 m =
280/0.025 = 11,200 m.

If you can get this depth you will have sufficient pressure differential to achieve
reverse osmosis with the inner freshwater stable at sea level.  Deeper water will
achieve actual flow.  Assume at this stage that the water inside the tube and
outside the tube are in thermal equilibrium (given the low flow rates likely, this
is reasonable) to discount temperature controlled density difference.

The system would work, if there is sufficient circulation (this is the energy input
bit) at depth to remove bypass water (the bit left behind after RO). In other words
the energy input is current flow which is ultimately solar and tidal. It is not
perpetual motion.  However, given the maintenance requirements to prevent fouling of
the membrane (11 km below surface kind of rules out snorkel, fins and a scrubbing
brush) and the difficulty of initial installation, the scheme is not exactly
practical.  Maybe you could use robots :-).  You also better hope the whole lot
doesn't get wiped out by a turbidity current after an earthquake.  Don't forget the
initial energy and financial input to manufacture the system.

Ultimately if your going to harness solar energy or tidal forces there are much more
efficient (less inefficient?) ways of doing it.

Theoretically possible? Yes.
Practical? Noooooo.

Spend your money on a decent coal/nuclear/solar/cold-fusion/rat-in-cage powered
generator, and run a commercial RO unit if you want fresh water.

Back to work
Robert Virtue

By the way OSMOSIS is  what you get when you turn off the power at an RO plant :-)