> As it turns out, doing that takes a lot of energy
The change in entropy between a batch of saline water and a batch of fresh water and enough saline water that its concentration don't change is about the same as letting that same fresh water fall for 200m and converting the resulting energy into heat (at 300K).
What means that desalination will take a lot of energy whatever method you use. There are distillation procedures close to perfect efficiency that wouldn't take much more energy than reverse osmosis; and of course, electrical separation is that one method with lots of promise but that stops due to material related problems every time it's tried. It just so happen that we know how to scale reverse osmosis up cheaply and reliably; but this looks like a feature of our technology and not anything intrinsic.
The change in entropy between a batch of saline water and a batch of fresh water and enough saline water that its concentration don't change is about the same as letting that same fresh water fall for 200m and converting the resulting energy into heat (at 300K).
What means that desalination will take a lot of energy whatever method you use. There are distillation procedures close to perfect efficiency that wouldn't take much more energy than reverse osmosis; and of course, electrical separation is that one method with lots of promise but that stops due to material related problems every time it's tried. It just so happen that we know how to scale reverse osmosis up cheaply and reliably; but this looks like a feature of our technology and not anything intrinsic.